WO2022227393A1 - Image photographing method and apparatus, electronic device, and computer readable storage medium - Google Patents

Abstract

The present disclosure relates to an image photographing method and apparatus, an electronic device, and a computer readable storage medium. The method comprises: obtaining a scene image corresponding to a photographing interface; recognizing the scene image, and determining a scene category corresponding to the scene image and a recommended human body posture corresponding to the scene category; performing human body key point detection on the scene image, and determining an object in the scene image and a real human body posture of the object; displaying the real human body posture and the recommended human body posture in the photographing interface to guide the object to perform posture adjustment according to the recommended human body posture; and when a photographing condition is satisfied, obtaining a photographing image of the object. According to the embodiments of the present disclosure, the image photographing effect can be effectively improved, such that a photographing image that a user is satisfied with can be obtained.

Description

Image capturing method and apparatus, electronic device and computer-readable storage medium

CROSS-REFERENCE TO RELATED APPLICATIONS

The present disclosure is based on the Chinese patent application with the application number of 202110467998.6, the application date of April 28, 2021, and the application title of "image capturing method and device, electronic equipment and storage medium", and claims the priority of the Chinese patent application, The entire contents of this Chinese patent application are hereby incorporated by reference into the present disclosure.

technical field

The present disclosure relates to the field of computer technologies, and in particular, to an image capturing method and apparatus, an electronic device, and a computer-readable storage medium.

Background technique

With the development of camera technology, users can take pictures by using the shooting function on mobile terminals such as mobile phones and tablet computers. When the user lacks photographing skills, or the user's limbs are not natural enough in front of the camera and the photographing posture is single, the photographed images may not satisfy the user, which reduces the photographing experience.

SUMMARY OF THE INVENTION

The embodiments of the present disclosure provide an image capturing technical solution.

According to an aspect of the present disclosure, an image capturing method is provided, including: acquiring a scene image corresponding to a shooting interface; recognizing the scene image, determining a scene category corresponding to the scene image and a scene category corresponding to the scene category the recommended human body posture; perform human key point detection on the scene image to determine the object in the scene image and the real human body posture of the object; by displaying the real human body posture and the recommended body posture in the shooting interface Human body posture, guide the object to perform posture adjustment according to the recommended human body posture; and obtain a photographed image of the object when the photographing conditions are met. In this way, the image shooting effect is effectively improved, the shooting experience of the user is improved, and the shooting image that satisfies the user can be obtained.

In a possible implementation manner, the recommended human posture includes a single-person posture, and the displaying the real human posture and the recommended human posture in the shooting interface includes: according to the object in the scene The position in the image, the real human body posture and the single-person posture are displayed in the shooting interface; or, according to the position of the object in the scene image, the real human body is displayed in the shooting interface. gesture, and the single-person gesture is displayed at the first designated area of the photographing interface. In this way, the object can be effectively guided to complete the shooting of the single-person posture, the shooting experience can be improved, and the recommended human posture can be displayed in a fixed manner and can be displayed with the object, which can meet different posture display requirements.

In a possible implementation manner, the recommended human body posture includes a multi-person combined posture, and the method further includes: according to relative positions between multiple objects in the scene image, and each of the multi-person combined postures The relative position between the gestures, and the corresponding gestures of each object are determined from the combined gestures of the multiple people; wherein, displaying the real human body gesture and the recommended human body gesture in the shooting interface includes: for any object , according to the position of the object in the scene image, display the real human body posture of the object and the corresponding posture of the object in the shooting interface; or, according to the position of the object in the scene image , the real human body posture of the object is displayed in the shooting interface, and the corresponding posture of the object is displayed in the second designated area of the shooting interface. In this way, multiple objects can be effectively guided to complete the combined pose of multiple people, and the shooting experience can be improved, and the recommended human pose can be displayed in a fixed manner and can be displayed with the object, which can meet different pose display requirements.

In a possible implementation manner, the method further includes: determining a key point pair between the first human body key point of the real human body posture and the second human body key point of the recommended human body posture; the first similarity of the key point pair, and from the first human body key points, determine a third human body key point whose first similarity is less than a first preset threshold; wherein, described in the shooting interface Displaying the real human body posture and the recommended human body posture includes: highlighting the region where the third human body key point is located, wherein the highlighting method includes at least one of highlighting, bolding, and changing color. In this way, the parts of the human body with low similarity among the displayed real human body postures can be highlighted, so as to more effectively guide the user to adjust the posture, so as to improve the shooting experience.

In a possible implementation manner, the method further includes: determining the real human body posture and the recommended human body according to a first similarity of key point pairs between the real human body posture and the recommended human body posture The second similarity between postures, the key point pair is determined according to the first human body key point of the real human body posture and the second human body key point of the recommended human body posture; the second similarity. In this way, the overall similarity between the real human body posture and the recommended human posture can be displayed, so as to effectively guide the user to adjust the posture according to the overall similarity, and improve the shooting experience.

In a possible implementation manner, the method further includes: inputting the scene image to an area recommendation network to obtain a recommended photographing area of the scene image, where the recommended photographing area is used to represent the scene image in the scene image A recommended photographing area, the area recommendation network is a neural network obtained by training the first sample set marking the photographing area; wherein, after obtaining the recommended photographing area of the scene image, the method further includes at least one of the following: In the shooting interface, an identifier is used to indicate the recommended photographing area; the recommended human posture is displayed at the recommended photographing area of the photographing interface, and the recommended photographing area is used to guide the object according to the It is recommended to adjust the position of the photo area. In this way, a photographing area with better visual effect can be recommended to the subject, and the subject can be effectively guided to adjust the position to improve the visual effect of the photographed image, thereby improving the user's photographing experience.

In a possible implementation manner, recognizing the scene image, determining a scene category corresponding to the scene image and a recommended human pose corresponding to the scene category, includes: performing a pose recommendation network on the scene image. Identify and determine the scene category corresponding to the scene image and the recommended human posture corresponding to the scene category, wherein the posture recommendation network is a neural network obtained by training the second sample set marked with the sample scene category and the sample human posture. network, the second sample set includes sample images uploaded by the object. In this way, the pose recommendation network can learn the human pose from the high-quality images, thereby effectively outputting high-quality recommended human poses based on the scene images.

In a possible implementation manner, the recommended human body postures include multiple ones, and the method further includes: displaying in the shooting interface a plurality of posture diagrams corresponding to the recommended human body postures; in response to the selection of the posture diagrams Operation, according to the selected posture map, determine and display the recommended human body posture corresponding to the selected posture map. In this way, it is convenient for the user to select different recommended human body postures, to satisfy different posture preferences, to improve the shooting experience, and to obtain a photographed image that satisfies the user.

In a possible implementation manner, the shooting conditions include: a second degree of similarity between the real human body posture and the recommended human body posture is greater than or equal to a second preset threshold; or, in the shooting interface Shooting controls are triggered.

In a possible implementation manner, the method further includes: processing the captured image to obtain a processed captured image, wherein the processing includes at least one of the following: performing processing on an object in the captured image Beauty processing, adding a filter to the captured image; the adding a filter to the captured image includes: adjusting the saturation, color temperature, color temperature, etc. of the captured image according to the recommended filter corresponding to the recommended human posture. at least one of brightness. In this way, the visual effect of the captured image can be improved, which is conducive to obtaining a captured image that satisfies the user, and improves the image capturing experience.

In a possible implementation manner, the method is applied to an artificial intelligence education platform, and the artificial intelligence education platform is connected with intelligent hardware; wherein, the artificial intelligence education platform is used to edit the project code for realizing the image capturing method , and send the project code to the intelligent hardware; the intelligent hardware is used to collect scene images and run the project code to obtain the operation result, and send the operation result to the artificial intelligence education platform to The operation result is displayed on the display interface of the artificial intelligence education platform. In this way, users can learn through interesting cases, and learn artificial intelligence algorithms such as human body gesture recognition and human body key point detection.

According to an aspect of the present disclosure, there is provided an image capturing apparatus, comprising: an acquisition part configured to acquire a scene image corresponding to a shooting interface; a recognition part configured to recognize the scene image and determine the scene The scene category corresponding to the image and the recommended human body posture corresponding to the scene category; the detection part is configured to perform human key point detection on the scene image, and determine the object in the scene image and the real human body posture of the object The display part is configured to guide the object to adjust the posture according to the recommended human body posture by displaying the real human body posture and the recommended human body posture in the shooting interface; the shooting part is configured to meet the requirements of shooting condition, a captured image of the object is obtained.

In a possible implementation manner, the recommended human posture includes a single-person posture, and the display part includes: a first display sub-part, configured to display the position of the object in the scene image according to the position of the object in the scene image. The real human body posture and the single-person posture are displayed in the shooting interface; or, the second display subsection is configured to display the object in the shooting interface according to the position of the object in the scene image. The real human body posture is displayed, and the single-person posture is displayed in the first designated area of the shooting interface.

In a possible implementation manner, the recommended human body posture includes a multi-person combined posture, and the apparatus further includes: a position determination part configured to The relative position between each posture in the multi-person combined posture, and the corresponding posture of each object is respectively determined from the multi-person combined posture; wherein, the display part includes: a third display sub-section, which is configured for any an object, according to the position of the object in the scene image, display the real human body posture of the object and the corresponding posture of the object in the shooting interface; or, the fourth display sub-section is configured to be based on The position of the object in the scene image, the real human body posture of the object is displayed in the shooting interface, and the corresponding posture of the object is displayed in the second designated area of the shooting interface.

In a possible implementation manner, the apparatus further includes: a first determining part configured to determine the distance between the first human body key point of the real human body posture and the second human body key point of the recommended human body posture The key point pair; the second determination part is configured to determine from the first human body key points that the first similarity is less than the first preset threshold by calculating the first similarity of the key point pair. The third human body key point; wherein, the display part, including: a highlight sub-section, is configured to highlight the region where the third human body key point is located, wherein the highlighting method includes highlighting, bolding, changing at least one of the colors.

In a possible implementation manner, the apparatus further includes: a third determination part, configured to determine the the second similarity between the real human body posture and the recommended human body posture, and the key point pair is determined according to the first human body key point of the real human body posture and the second human body key point of the recommended human body posture; The similarity display part is configured to display the second similarity in the shooting interface.

In a possible implementation manner, the apparatus further includes: an area determination part, configured to input the scene image into an area recommendation network to obtain a recommended photographing area of the scene image, where the recommended photographing area is used for Representing the area recommended to be photographed in the scene image, and the area recommendation network is a neural network trained by marking the first sample set of the photographing area; wherein, the device further includes: an area display part, configured to In the shooting interface, an identifier is used to indicate the recommended photographing area; or, the recommended human posture is displayed in the recommended photographing area of the photographing interface; the recommended photographing area is used to guide the object according to the recommended photographing area. Adjust the position of the recommended photo area as described above.

In a possible implementation manner, the area display part is further configured to use an identifier to indicate the recommended photographing area in the photographing interface; and display the recommended photographing area on the photographing interface. The recommended human posture is described.

In a possible implementation manner, the identifying part is further configured to: identify the scene image through a gesture recommendation network, and determine a scene category corresponding to the scene image and a recommended human body corresponding to the scene category pose, wherein the pose recommendation network is a neural network obtained by training a second sample set marked with sample scene categories and sample human poses, and the second sample set includes sample images uploaded by objects.

In a possible implementation manner, the recommended human body postures include a plurality of postures, and the apparatus further includes: a posture map display part configured to display the posture maps corresponding to the plurality of recommended human body postures in the shooting interface; selecting The part is configured to, in response to a selection operation on the posture map, determine and display a recommended human body posture corresponding to the selected posture map according to the selected posture map.

In a possible implementation manner, the shooting conditions include: a second degree of similarity between the real human body posture and the recommended human body posture is greater than or equal to a second preset threshold; or, in the shooting interface Shooting controls are triggered.

In a possible implementation manner, the apparatus further includes: a processing part configured to process the captured image to obtain a processed captured image, wherein the processing includes at least one of the following: Performing beautification processing on the object in the captured image, and adding a filter to the captured image; the adding a filter to the captured image includes: adjusting the captured image according to the recommended filter corresponding to the recommended human posture at least one of saturation, color temperature, and brightness.

In a possible implementation, the device is applied to an artificial intelligence education platform, and the artificial intelligence education platform is connected to intelligent hardware; wherein, the artificial intelligence education platform is used to edit and implement the project code of the image capturing device , and send the project code to the intelligent hardware; the intelligent hardware is used to collect scene images and run the project code to obtain the operation result, and send the operation result to the artificial intelligence education platform to The operation result is displayed on the display interface of the artificial intelligence education platform.

According to an aspect of the present disclosure, there is provided an electronic device, comprising: a processor; a memory configured to store instructions executable by the processor; wherein the processor is configured to invoke the instructions stored in the memory to execute the above method.

According to an aspect of the present disclosure, there is provided a computer-readable storage medium having computer program instructions stored thereon, the computer program instructions implementing the above method when executed by a processor.

According to an aspect of the present disclosure, there is provided a computer program comprising computer-readable code, which, when the computer-readable code is executed in an electronic device, implements the above method when executed by a processor in the electronic device.

In the embodiment of the present disclosure, the recommended human body pose corresponding to the scene category can be recommended to the user according to the scene category of the scene image, so that the recommended human body pose matches the scene image, and by displaying the real human body pose and The recommended human posture can guide the user to adjust the posture according to the displayed recommended human posture or facilitate the user to guide the subject to adjust the posture according to the displayed recommended human posture. Satisfied taking pictures.

It is to be understood that the foregoing general description and the following detailed description are exemplary and explanatory only and are not restrictive of the present disclosure. Other features and aspects of the present disclosure will become apparent from the following detailed description of exemplary embodiments with reference to the accompanying drawings.

Description of drawings

The accompanying drawings, which are incorporated into and constitute a part of this specification, illustrate embodiments consistent with the present disclosure, and together with the description, serve to explain the technical solutions of the present disclosure.

FIG. 1 shows a flowchart of an image capturing method according to an embodiment of the present disclosure.

FIG. 2 shows a schematic diagram of a photographing interface according to an embodiment of the present disclosure.

FIG. 3 shows a schematic diagram of a multi-person combined gesture according to an embodiment of the present disclosure.

FIG. 4 shows a schematic diagram of a real human body pose according to an embodiment of the present disclosure.

FIG. 5 shows a schematic diagram of a photographing interface according to an embodiment of the present disclosure.

FIG. 6 shows a schematic diagram of a scene image according to an embodiment of the present disclosure.

FIG. 7 shows a schematic diagram of a photographing interface according to an embodiment of the present disclosure.

FIG. 8 shows a schematic diagram of a photographing interface according to an embodiment of the present disclosure.

FIG. 9 shows a schematic diagram of a photographing interface according to an embodiment of the present disclosure.

FIG. 10 shows a schematic diagram of a photographing interface according to an embodiment of the present disclosure.

11a and 11b illustrate schematic diagrams of a photographing interface according to an embodiment of the present disclosure.

FIG. 12 shows a schematic diagram of an application of an image capturing method according to an embodiment of the present disclosure.

FIG. 13 shows a block diagram of an image capturing apparatus according to an embodiment of the present disclosure.

FIG. 14 shows a block diagram of an electronic device according to an embodiment of the present disclosure.

Detailed ways

Various exemplary embodiments, features and aspects of the present disclosure will be described in detail below with reference to the accompanying drawings. The same reference numbers in the figures denote elements that have the same or similar functions. While various aspects of the embodiments are shown in the drawings, the drawings are not necessarily drawn to scale unless otherwise indicated.

The word "exemplary" is used exclusively herein to mean "serving as an example, embodiment, or illustration." Any embodiment described herein as "exemplary" is not necessarily to be construed as preferred or advantageous over other embodiments.

The term "and/or" in this article is only an association relationship to describe the associated objects, indicating that there can be three kinds of relationships, for example, A and/or B, it can mean that A exists alone, A and B exist at the same time, and A and B exist independently B these three cases. In addition, the term "at least one" herein refers to any combination of any one of the plurality or at least two of the plurality, for example, including at least one of A, B, and C, and may mean including from A, B, and C. Any one or more elements selected from the set of B and C.

It should be understood that the terms "first", "second" and "third" in the claims, description and drawings of the present disclosure are used for distinguishing descriptions, rather than for describing a specific order, nor can they be construed as indications Or imply relative importance. The terms "comprising" and "comprising" as used in the specification and claims of the present disclosure indicate the presence of the described feature, integer, step, operation, element and/or component, but do not exclude one or more other features, integers , step, operation, element, component and/or the presence or addition of a collection thereof.

In addition, in order to better illustrate the present disclosure, numerous specific details are set forth in the following detailed description. It will be understood by those skilled in the art that the present disclosure may be practiced without certain specific details. In some instances, methods, means, components and circuits well known to those skilled in the art have not been described in detail so as not to obscure the subject matter of the present disclosure.

FIG. 1 shows a flowchart of an image capturing method according to an embodiment of the present disclosure. As shown in FIG. 1 , the image capturing method includes:

In step S11, obtain a scene image corresponding to the shooting interface;

In step S12, the scene image is identified, and the scene category corresponding to the scene image and the recommended human posture corresponding to the scene category are determined;

In step S13, the human body key point detection is performed on the scene image, and the object in the scene image and the real human body posture of the object are determined;

In step S14, the real human body posture and the recommended human body posture are displayed in the shooting interface, so as to guide the object to adjust the posture according to the recommended human body posture;

In step S15, when the photographing conditions are satisfied, a photographed image of the object is obtained.

In a possible implementation manner, the image capturing method may be performed by a terminal device, and the terminal device may be a user equipment (User Equipment, UE), a mobile device, a user terminal, a terminal, a cellular phone, a cordless phone, or a personal digital assistant (Personal Digital Assistant, PDA), handheld device, computing device, vehicle-mounted device, wearable device, etc., the method can be implemented by the processor calling the computer-readable instructions stored in the memory. This embodiment of the present disclosure does not limit the type of the terminal device.

In a possible implementation manner, an application (Application, APP) implementing the image capturing method may be deployed in the terminal device, so that the user can directly use the application to capture images; As a shooting mode, it is integrated into the shooting function of the terminal device, so that when the user uses the shooting function of the terminal device, the user can select the shooting mode that realizes the image shooting method to shoot the image, which is not made in this embodiment of the present disclosure. limit.

It should be understood that, after starting an application for implementing the image capturing method in the present disclosure, or selecting a capturing mode for implementing the image capturing method in the present disclosure, an image capturing device (such as a camera) mounted on the terminal device may be used, or an image capturing device (such as a camera) assembled on the terminal device or a The image acquisition device connected to the device collects the scene image of the actual scene in real time and displays it on the shooting interface of the terminal device. In step S11, a scene image corresponding to the shooting interface is acquired, that is, a scene image acquired by an image acquisition device is acquired.

FIG. 2 shows a schematic diagram of a shooting interface according to an embodiment of the present disclosure. As shown in FIG. 2 , a scene image can be displayed in the shooting interface, and a shooting control 22 for triggering shooting can be provided in the shooting interface, which is used to check the shooting interface. A viewing control 21 for an image, a triggering control 23 for triggering recognition of a scene image, and the like.

It should be noted that the shooting interface shown in FIG. 2 is an implementation provided by the embodiment of the present disclosure. It should be understood that the present disclosure should not be limited to this, and those skilled in the art can set the functions in the shooting interface according to actual needs. control, which is not limited by this embodiment of the present disclosure.

In a possible implementation manner, the image capturing method may also be performed by a server, and the server may be an independent physical server, a server cluster or a distributed system composed of multiple physical servers, or a cloud-provided server. Services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, CDN, and cloud servers for basic cloud computing services such as big data and artificial intelligence platforms, etc. The disclosed embodiments do not limit the type of server.

In the case where the image capturing method is executed by the server, in step S11, a scene image corresponding to the shooting interface is acquired, that is, the server receives the scene image sent by the terminal, wherein the scene image is collected by the terminal through the image capturing device image.

In a possible implementation manner, in step S12, the recognition of the scene image can be triggered by triggering the trigger control 23 as shown in FIG. After the shooting mode of the image shooting method, the scene image recognition is automatically triggered, which is not limited in this embodiment of the present disclosure.

In a possible implementation manner, in step S12, recognizing the scene image, determining the scene category corresponding to the scene image and the recommended human pose corresponding to the scene category, may include: recognizing the scene image, and obtaining the scene image in the scene image. at least one of the scene feature and the object feature of the image; according to at least one of the identified scene feature and the object feature, determine the scene type corresponding to the scene image; according to the correspondence between the preset scene type and the recommended human posture , and determine the recommended human pose corresponding to the scene type.

The scene features can be used to indicate the scene type of the actual scene where the object is located. For example, the scene features of the grassland can include grass, cattle and sheep, etc.; the scene features of the indoor can include walls, windows, etc.; if the identified scene features include Grassland and cattle and sheep, the scene type corresponding to the obtained scene image is grassland.

In a possible implementation manner, the object features may include at least one of the following: clothing features, physical features, gender features, and the like. Among them, the clothing feature can be used to indicate the clothing of the object, the body feature can be used to indicate the body shape of the object (for example, tall and small), and the gender feature can be used to indicate the gender of the object. In this way, the scene type can be determined in combination with the scene feature and the object feature, so as to obtain a more accurate recommended human pose.

For example, if the scene feature indicates that the actual scene is indoor, the object feature indicates that the object is wearing "sweater, trousers, handbag", and the object is a small girl, the scene category can be "indoor-small-girl-wearing sweater, trousers" , handbag”, and then determine the corresponding recommended human pose according to this scene category.

The scene image can be recognized by a pre-trained scene recognition network to obtain scene features and/or object features in the scene image, and the embodiments of the present disclosure do not limit the network type, network result, and training method of the scene recognition network.

In a possible implementation manner, the identified scene features and object features can also be displayed in the shooting interface, for example, the scene features and object features can be marked by means of tags, so as to improve the interest of image shooting.

It should be understood that different scene types may correspond to different recommended human postures, and the corresponding relationship between scene types and recommended human postures can be preset. The recommended human pose corresponding to the scene category.

In a possible implementation manner, in step S12, a pre-trained gesture recommendation network may also be used to identify the scene image, and determine the scene category corresponding to the scene image and the recommended human pose corresponding to the scene category. The embodiments of the present disclosure do not limit the network type, network structure, training method, etc. of the gesture recommendation network.

In a possible implementation, a self-supervised training method can be used to train the pose recommendation network to learn the recommended human poses under different scene types, or, in other words, to learn the correspondence between the scene types and the recommended human poses, so as to use The pose recommendation network directly obtains the recommended human pose based on the scene image. In this way, compared with the above-mentioned method of determining the recommended human posture according to the corresponding relationship between the preset scene type and the recommended human posture, more abundant scene types and recommended human postures can be obtained intelligently, and more accurate It recommends the recommended human pose corresponding to the scene image to the user.

In a possible implementation manner, the recommended human body posture may include the human body posture learned from the sample images by the above-mentioned posture recommendation network, for example, may also include the human body posture manually drawn by a user (eg, a technician), to which the embodiments of the present disclosure No restrictions apply. It should be understood that the recommended human posture may be represented in the form of a human skeleton (eg, a skeleton obtained by connecting human joint points), or in the form of a human body outline, which is not limited to this embodiment of the present disclosure.

Among them, the sample image is a high-quality image with better shooting effect. Learning the human body posture from the high-quality image and recommending it to the user can guide the object to approach the human body posture in the high-quality image, so as to obtain a better shooting effect and improve the user's shooting experience.

In a possible implementation manner, the recommended human posture may include a single-person posture (for example, a single-person comparison), and may also include a multi-person combined posture (for example, a two-person combination). It should be understood that the recommended human gesture corresponding to the scene category may include at least one single-person gesture and/or at least one multi-person combined gesture. FIG. 3 shows a schematic diagram of a multi-person combined posture according to an embodiment of the present disclosure. As shown in FIG. 3 , the posture n and the posture m together form a posture of “comparing hearts”.

In a possible implementation manner, the scene types may, for example, at least include: landscape, indoor, outdoor, etc.; wherein, the scenery can be divided into beach, landscape, grassland, lake, etc., and the indoor can be divided into For office, conference room, home, etc., outdoor category can be divided into: street view category, playground category, park category, etc.

It should be understood that the classification of scene types may not be limited to this. In fact, those skilled in the art can preset various scene types according to actual needs, or can also self-learn scene types through the above gesture recommendation network. Examples are not limited.

In a possible implementation manner, in step S13, any known human body key point detection method may be used, such as using a human body key point detection network to perform human key point detection on the scene image, which is not made in this embodiment of the present disclosure. limit. The embodiments of the present disclosure do not limit the network type, network structure, and training method of the human body key point detection network.

In a possible implementation manner, the detection of human body key points on the scene image may include: extracting the key points of human body joint parts of objects in the scene image (for example, the human body key points of 20 joint parts), wherein the key points of the human body joint parts are extracted. The number and position of the points may be determined according to actual requirements, which are not limited in this embodiment of the present disclosure.

In a possible implementation manner, the object in the scene image and the coordinate values of the key points of multiple human joint parts of the object in the scene image can be determined according to the detected key points of the human body joints; value, and connect the key points of multiple human joint parts according to the structure of the human body to obtain the human skeleton of the object, that is, to obtain the real human body posture of the object. FIG. 4 shows a schematic diagram of a real human body posture according to an embodiment of the present disclosure. As shown in FIG. 4 , 20 key points of joint parts of the human body are connected to obtain the real human body posture of the object.

In a possible implementation manner, in step S13, performing human body key point detection on the scene image may also include: extracting contour key points on the human body contour of the object in the scene image, obtaining the human body contour of the object, and then using the human body The contour represents the real human pose of the object. The manner in which the posture of the human body is represented may be determined according to actual needs, which is not limited by the embodiment of the present disclosure.

It should be noted that, the above-mentioned steps S12 and S13 may be performed simultaneously; or step S12 may be performed first, and then step S13 may be performed; or step S13 may be performed first, and then step S12 may be performed. Specifically, it may be set according to factors such as the processing capability of the terminal device, the resource occupancy of the terminal device, and the time delay limit in the application process, which is not limited in this embodiment of the present disclosure.

In a possible implementation manner, in step S14, displaying the real human body posture and the recommended human body posture in the shooting interface may include: displaying the human body skeleton corresponding to the recommended human body posture and the human body skeleton corresponding to the real human body posture in the shooting interface ; or, displaying the human body contour corresponding to the recommended human body posture and the human body contour corresponding to the real human body posture in the shooting interface. In other words, the form of human skeleton or human outline can be used in the shooting interface to display the real human body posture and the recommended human body posture. In this way, the user can be guided to adjust the posture according to the displayed recommended human body posture, so as to achieve a state with a high similarity to the recommended human body posture, thereby obtaining a better shooting effect.

FIG. 5 shows a schematic diagram of a shooting interface according to an embodiment of the present disclosure. As shown in FIG. 5 , the real human body posture and the recommended human body posture of the object can be displayed based on the form of the human skeleton; in a possible implementation manner , the real human body posture and the recommended human body posture are displayed in different display forms; for example, different colors can be used to represent the real human body posture and the recommended human body posture of the object, for example, the recommended human body posture is represented by yellow, and the recommended human body posture is represented by green. Real human pose. In a possible implementation, as shown in FIG. 5 , guide information can also be displayed in the shooting interface, such as displaying the similarity between the real human body posture and the recommended human posture: 30%, and displaying the information that guides the user to adjust the posture. The prompt "Try another action? Try to be consistent with the yellow line".

It should be understood that the real human body posture and the representation of the recommended human body posture shown in FIG. 5 are an implementation manner provided by the embodiments of the present disclosure, and the present disclosure should not be limited to this. It is required to set the display form of the real human body posture and the recommended human body posture, such as color, line thickness, line type, area transparency, etc., as well as setting the content of the guidance information, which is not limited by this embodiment of the present disclosure.

In a possible implementation manner, in step S15, when the image capturing method is executed by the terminal device, the capturing conditions may include that the capturing control in the capturing interface is triggered. It can be understood that the photographer clicks the 2. The shooting controls in the shooting interface shown in FIG. 4 and FIG. 5 trigger the shooting operation, obtain the shooting image of the object and save the local album for viewing.

In the embodiment of the present disclosure, the recommended human body pose corresponding to the scene category can be recommended to the user according to the scene category of the scene image, so that the recommended human body pose matches the scene image, and by displaying the real human body pose and The recommended human posture can guide the user to adjust the posture according to the displayed recommended human posture or facilitate the user to guide the subject to adjust the posture according to the displayed recommended human posture. Satisfied taking pictures.

As described above, the recommended human posture may include a single-person posture. In a possible implementation manner, in step S14, the real human posture and the recommended human posture are displayed on the shooting interface, including:

According to the position of the object in the scene image, the real human pose and the single-person pose are displayed in the shooting interface; or,

According to the position of the object in the scene image, the real human body pose is displayed in the shooting interface, and the single-person pose is displayed at the first designated area of the shooting interface.

It should be understood that one or more objects may be included in the scene image. The above-mentioned human body key point detection on the scene image may include: performing human body detection on the scene image to obtain the object in the scene image and the human body area of the object; and then performing human body key point detection on the object in the human body area. It should be understood that the human body detection can obtain information such as whether an object is included in the scene image, the number of objects included, and the position of the object in the scene image (ie, the position of the human body region in the scene image).

For the case where the recommended human posture is a single-person posture, if the scene image contains an object, the real human posture of the object can be displayed directly according to the position of the object; if the scene image contains multiple objects, the preset The selection rule is to select an object from multiple objects, and display the real human body posture of the selected object according to the position of the selected object.

In a possible implementation manner, the preset selection rule may include, for example, selecting an object in the middle of a scene image, or selecting an object closest to the terminal device, etc., which are not limited in this embodiment of the present disclosure. The distance between the object and the terminal device may be obtained by using a technique known in the art, for example, a time of flight (TOF) method may be used, which is not limited in this embodiment of the present disclosure.

As described above, the real human pose can be represented and the human pose can be recommended based on the form of human skeleton or human outline. In a possible implementation manner, the position of the object in the scene image may include: the position coordinates of the key points of the human body on the human skeleton of the object in the scene image or the human body contour; may also include the human body region of the object in the scene image The position of the object, wherein the human body area of the object can be obtained through the above-mentioned human body detection.

It should be understood that, through the detection of the human body key points, the position coordinates of the human body key points can be obtained, and then the real human body posture can be displayed in the shooting interface according to the position coordinates of the human body key points. In this way, the real human body posture in the photographing interface can be displayed following the position of the object.

In a possible implementation manner, for the recommended human posture, it can be set to follow the position of the object to display, that is, to display the recommended human posture according to the position of the object in the scene image; it can also be set to the first specified in the shooting interface. The area displays the recommended human body posture, that is, the recommended human body posture is fixedly displayed in a preset area of the shooting interface.

In a possible implementation manner, according to the position of the object in the scene image, displaying the single-person posture in the shooting interface may include: according to the position of the human body region of the object in the scene image, displaying on the human body region of the shooting interface Solo pose. In this way, it is relatively convenient to realize that the recommended human body posture is displayed according to the position of the object, and it is convenient to guide the object to compare the posture of a single person with the posture of a real human body for posture adjustment.

In a possible implementation manner, according to the position of the object in the scene image, the single-person posture is displayed in the shooting interface, and may further include: according to the position coordinates of any key point of the human body (such as the key point of the neck) on the real human body posture to display the single-person pose in the shooting interface. This method can be understood as making any key point of the human body in the single-person posture coincide with the key point of the human body corresponding to the real human body posture. For example, the key points of the neck on the two postures can be set to overlap, and the single-person posture can be displayed on this basis. , so as to realize the position display of the single-person posture following the object. In this way, the single-person posture can be displayed more accurately following the position of the object, which is convenient for the user to compare the single-person posture and the real human body posture.

It should be noted that the above manners of displaying the pose of a single person in the shooting interface are some implementation manners provided by the embodiments of the present disclosure. In fact, the present disclosure should not be limited, as long as the single person is displayed according to the position of the object in the scene image The implementation manner of the gesture is all within the protection scope of the present disclosure.

In a possible implementation manner, the first designated area in the shooting interface can be set according to actual needs, for example, it can be the middle area, the left area, the right area, the four vertex areas, etc. of the shooting interface. The embodiments of the present disclosure are not limited. By displaying the single-person posture at the first designated area of the photographing interface, the recommended human body posture can be fixedly displayed in the photographing interface, thereby saving computing resources required for drawing the recommended human body posture.

The size of the range of the first designated area and the display size of the single-person posture in the first designated area may be set according to actual requirements, which are not limited in this embodiment of the present disclosure.

In the embodiment of the present disclosure, the object can be effectively guided to complete the shooting of the single-person posture, and the shooting experience can be improved, and the recommended human posture can be displayed fixedly and can be displayed with the object, which can meet different posture display requirements.

As mentioned above, the recommended human body posture includes the combined posture of multiple people. In a possible implementation manner, the method further includes:

According to the relative positions between the multiple objects in the scene image and the relative positions between the respective gestures in the multiple-person combined gestures, the corresponding gestures of each object are respectively determined from the multiple-person combined gestures. In this way, the corresponding postures of each object in the scene image in the multi-person combined posture can be effectively determined, so that it is convenient to guide each object to complete a plurality of combined postures.

As mentioned above, one or more objects may be included in the scene image. Performing human body key point detection on the scene image may include: performing human body detection on the scene image to obtain the object in the scene image and the human body area of the object; and then performing human key point detection on the object in the human body area. It should be understood that the human body detection can obtain information such as whether an object is included in the scene image, the number of objects included, and the position of the object in the scene image (ie, the position of the human body region in the scene image).

For the case where the recommended human pose is a multi-person combined pose, in a possible implementation manner, if the number of objects in the scene image is less than the number of objects required to realize the multi-person combined pose, for example, through voice or text, etc. Prompt the user that the number of people is not enough to achieve a multi-person combined posture, thereby guiding the increase of the number of objects to be photographed; or select a human posture that matches the number of objects in the scene image from the multi-person combined posture, and determine the corresponding posture of the object based on the selected human posture .

Among them, the human body posture that matches the number of objects in the scene image is selected from the combined postures of multiple people. For example, the human body posture that matches the number of objects can be randomly selected, or the human body posture that matches the number of objects can be selected according to a preset selection strategy. The selection strategy may include, for example, selection in order from left to right and from top to bottom, which is not limited to this embodiment of the present disclosure.

It should be understood that if the selected human body posture is one, the selected human body posture can be directly used as the corresponding posture of the object; if there are multiple selected human body postures, according to the relative positions between the multiple selected human body postures, Determine the corresponding pose of each object.

If the number of objects in the scene image is greater than or equal to the number of objects required to realize the multi-person combined gesture, the relative positions of the multiple objects in the scene image and the relative positions of the various gestures in the multi-person combined gesture can be determined. The corresponding poses of each object are determined from the combined poses of multiple people. Wherein, the relative position may include, for example, a front-rear position, a left-right position, an up-down position, and the like, which is not limited by the embodiment of the present disclosure.

For example, FIG. 6 shows a schematic diagram of a scene image according to an embodiment of the present disclosure. As shown in FIG. 6 , the scene image includes object A, object B, and object C, and the current recommended human pose is, for example, the pose of a group of people shown in FIG. 3 (ie, the "comparison" pose). Then, according to the relative positions of object A, object B, and object C, as well as the relative positions of pose m and pose n in the combined pose of multiple people, it can be obtained that object A and object B correspond to pose m and pose n respectively, and also That is, it is determined that the corresponding posture of object A is "pose m", and the corresponding posture of object B is "pose n". It can be understood that the relative position between object A and object B is more suitable for completing the multi-person shown in FIG. Combining gestures.

In the case where the recommended human body posture is a multi-person combined posture, and the corresponding posture of each object is determined from the multi-person combined posture, in a possible implementation manner, in step S14, a real human body is displayed in the shooting interface Posture and recommended human posture, including:

For any object, according to the position of the object in the scene image, the real human body posture of the object and the corresponding posture of the object are displayed in the shooting interface; or, according to the position of the object in the scene image, the real human body of the object is displayed in the shooting interface. gesture, and the corresponding gesture of the object is displayed in the second designated area of the shooting interface.

Wherein, for any object among the multiple objects, according to the position of the object in the scene image, the real human body posture of the object is displayed in the shooting interface, and reference may be made to the content disclosed in the above embodiments of the present disclosure.

In a possible implementation manner, for any object among the multiple objects, according to the position of the object in the scene image, the corresponding posture of the object is displayed in the shooting interface, which can be the same as the method according to the object in the above-mentioned embodiment of the present disclosure. The implementation manner of displaying the single-person posture in the shooting interface at the position in the scene image is the same, which is not limited by the embodiment of the present disclosure. In this way, for each of the multiple objects, the corresponding posture of each object can be displayed following the position of each object, so as to facilitate guiding each object to compare the corresponding posture with the real human body posture for posture adjustment.

In a possible implementation manner, the second designated area in the shooting interface can be set according to actual needs, for example, it can be the middle area, left area, right area, four vertex areas, etc. of the shooting interface. The embodiments of the present disclosure are not limited. By displaying the corresponding posture of the object in the second designated area of the photographing interface, the recommended human body posture can be fixedly displayed in the photographing interface, thereby saving computing resources required for drawing the recommended human body posture.

It should be understood that the first designated area and the second designated area may be different or the same; the second designated area may include multiple areas, and each gesture in the combined gesture of multiple people may be displayed in the multiple areas; the second designated area also It may include an area in which the multi-person combined gesture is displayed, which is not limited in this embodiment of the present disclosure. FIG. 7 shows a schematic diagram of a shooting interface according to an embodiment of the present disclosure. As shown in FIG. 7 , a combined pose of multiple people can be displayed at a second designated area, and according to the position of each object in the scene image, it can be displayed in the shooting interface. The real human pose of each object.

The size of the range of the second designated area and the display size of the corresponding posture in the second designated area may be set according to actual needs, which is not limited by the embodiment of the present disclosure.

In the embodiment of the present disclosure, multiple objects can be effectively guided to jointly complete a multi-person combined pose, improving the shooting experience, and the recommended human pose can be displayed in a fixed manner and can be displayed with the object, which can meet different pose display requirements.

Considering that when the object adjusts the posture, there may be some human body postures that are close to the recommended human posture, while the postures of some human body parts are not close to the recommended human posture, for example, the posture of the arms is close to the recommended human posture (also That is, the similarity is high), and the posture of the legs is not close to the recommended human posture (that is, the similarity is low).

In order to guide the user to adjust the posture more effectively, the body parts with low similarity among the displayed real body postures can be highlighted. In a possible implementation, the method further includes:

Determine the key point pair between the first human body key point of the real human body posture and the second human body key point of the recommended human body posture; determine the first human body key point from the first human body key point by calculating the first similarity of the key point pair a third human body key point whose similarity is less than the first preset threshold;

Wherein, in step S13, the real human body posture and the recommended human body posture are displayed in the shooting interface, including:

The region where the third human body key point is located is highlighted, and the highlighting method includes at least one of highlighting, bolding, and changing color.

In a possible implementation manner, in order to facilitate the comparison between the real human body posture and the recommended human body posture, the first human body key point of the real human body posture and the second human body key point of the recommended human body posture may correspond. For example, the first human body key point of the real human body posture One human body key point includes 20 human body joint key points, and the second human body key point for the recommended human posture also includes 20 human body joint key points correspondingly. Based on this, key point pairs can be determined from the first human body key point of the real human body posture and the second human body key point of the recommended human body posture, for example, the left shoulder key point in the first human body key point, and the second human body key point. The left shoulder key point in the points is a pair of key point pairs, 20 key points of human body joints, and 20 key point pairs can be determined.

In a possible implementation manner, the first human body key point of the real human body posture may carry a mark, and the second human body key point of the recommended human body posture may also carry a mark. It should be understood that the marks can be used to indicate key points on different joint parts or different contour positions of the human body, and through the same/similar marks on the two postures, it is easy to determine the first human key point of the real human body posture, which is the same as that of the human body. Keypoint pairs between the second human keypoints for the recommended human pose.

It should be noted that the above method of determining key point pairs through identification is an implementation method disclosed in the embodiments of the present disclosure. In fact, those skilled in the art can use any known method to determine the first human body key point and the first human body key point. The key point pair between two human body key points is not limited in this embodiment of the present disclosure.

In a possible implementation manner, the first similarity of a keypoint pair can be represented by the distance (such as Euclidean distance, cosine distance, etc.) between two keypoints in the keypoint pair, or by the deviation of the distance, This embodiment of the present disclosure is not limited. The first similarity may be set to be negatively correlated with the distance or the deviation from the distance, that is, the smaller the distance or the smaller the deviation of the distance, the higher the first similarity. It should be understood that the number of key point pairs is consistent with the number of first degrees of similarity, that is, each key point pair corresponds to its respective first degree of similarity.

Among them, the deviation of the distance can be understood as the difference between the distance of any key point pair and the average value of the distances of all key point pairs. For example, the average value of the distances of 20 key point pairs is X, and the key point The distance to a is x, then the deviation of the distance of the key point to a is "x-X", that is, "x-X" can be used to represent the first similarity of the key point to a.

In a possible implementation manner, according to the display manner of the recommended human body posture: following the object display or fixed display, choose to use the distance or the deviation of the distance to represent the first similarity. For example, when the recommended human posture is displayed following the object, the distance can be used to represent the first similarity; when the recommended human posture is fixedly displayed in the photographing interface, the deviation of the distance can be used to represent the first similarity.

It should be understood that in the case where the recommended human pose is fixed and displayed in the shooting interface, the position of the object in the scene image will change, and the distance may not be able to accurately characterize the first similarity of the key point pair. This is because The long distance of any key point pair does not mean that the key point pair is not similar, it may be that the overall distance between the real human body posture and the recommended human body posture is far away. first similarity.

In a possible implementation manner, the first preset threshold may be determined according to an actual requirement, a calculation method of the first similarity, and the like, which is not limited in this embodiment of the present disclosure. The first similarity is less than the first preset threshold, which means that the similarity of the key point pair corresponding to the first similarity is low, that is, the similarity between the partial posture of the human body part represented by the key point pair and the recommended human posture On the contrary, if the first similarity degree is greater than or equal to the first preset threshold, it may mean that the similarity degree of the key point pair corresponding to the first similarity degree is high, that is, the partial posture of the human body part represented by the key point pair is the same as It is recommended that the human pose has a high similarity.

In a possible implementation manner, determining a third human body key point whose first similarity is less than a first preset threshold from the first human body key points by calculating the first similarity of the key point pair, may include: by Calculate the first similarity of the key point pair, determine the target key point pair whose first similarity is less than the first preset threshold from the key point pair, and use the first human body key point in the target key point pair as the third human body key point , that is, the third human body key point whose first similarity is less than the first preset threshold is determined from the first human body key point.

In a possible implementation manner, highlighting the area where the third human body key point is located may be understood as highlighting the area occupied by the third human body key point in the shooting interface. For example, in the real human body pose shown in FIG. 4 , a solid circle is used to represent the joint parts of the human body, then the area occupied by the solid circle may be the area where the third human body key point is located.

It should be understood that the highlighting manner includes at least one of highlighting, bolding, and changing color, which are some implementation manners provided by the embodiments of the present disclosure. In fact, those skilled in the art can design different highlighting according to actual needs. The manner of display is not limited to this embodiment of the present disclosure. The color is changed, for example, the third human body key point in the real human body posture displayed in green is changed to red, which is not limited in this embodiment of the present disclosure.

FIG. 8 shows a schematic diagram of a photographing interface according to an embodiment of the present disclosure. As shown in FIG. 8 , the key points 80 of the knee joint parts with lower similarity may be represented in a bold manner. The connection lines 81 and 82 corresponding to the key points 80 of the joint parts (that is, the connection lines representing the legs) are bolded, which is not limited by the embodiment of the present disclosure.

In the embodiment of the present disclosure, the parts of the human body with low similarity among the displayed real human body postures can be highlighted, so as to more effectively guide the user to adjust the posture, so as to improve the shooting experience.

As described above, the similarity between the real human body posture and the recommended human body posture can be displayed in the shooting interface to guide the subject to adjust the posture. In a possible implementation, the method further includes:

Determine the second similarity between the real human body posture and the recommended human body posture according to the first similarity of the key point pair between the real human body posture and the recommended human body posture, and the key point pair is the first human body key point according to the real human body posture Determined with the second human key point of the recommended human posture;

The second similarity is displayed in the shooting interface.

For the way of determining the key point pair, reference may be made to the above embodiments of the present disclosure.

As described above, the first similarity of the keypoint pair can be characterized by the distance of the keypoint pair or the deviation of the distance. In a possible implementation manner, determining the second similarity between the real human body posture and the recommended human posture according to the first similarity of the key point pairs may include: according to the accumulated value of the distances of all key point pairs, the average value, variance, or standard deviation to determine the second similarity, which is not limited in this embodiment of the present disclosure.

Among them, the accumulated value, average value, variance or standard deviation of the distance may be negatively correlated with the second similarity, that is, the smaller the accumulated value, mean, variance or standard deviation of the distance, etc., the higher the second similarity is. On the contrary, the larger the cumulative value, mean, variance or standard deviation of the distance, the lower the second similarity. It should be understood that the second similarity may reflect the overall similarity between the real human body posture and the recommended human body posture.

In a possible implementation manner, in order for the object to understand the level of the second similarity, or to understand the level of the similarity, the second similarity may be displayed in the shooting interface in the form of a percentage. Based on this, the cumulative value, average value, variance or standard deviation of the distance can be set, and there is a mapping relationship with the second similarity, so that according to the mapping relationship, the cumulative value, average value, variance or standard deviation of the distance can be mapped to The second similarity in percentage form.

Wherein, those skilled in the art may use any known manner to set the mapping relationship between the accumulated value, average value, variance or standard deviation of the distance and the second similarity, which is not limited in this embodiment of the present disclosure.

In a possible implementation manner, the second degree of similarity is displayed in the photographing interface, for example, the second degree of similarity may be displayed in the manner shown in FIG. 5 and FIG. 8 . It should be understood that a person skilled in the art can design a display manner of the second similarity according to actual requirements, which is not limited by this embodiment of the present disclosure.

In a possible implementation manner, the second similarity between the real human body posture and the recommended human body posture can also be obtained through a neural network, that is, the first human body key point corresponding to the real human body posture and the recommended human body posture can be obtained. The second human body key point is input into the neural network, and the second similarity is output. The first human body key point and the second human body key point can be input into the neural network in the form of a vector or a matrix. The embodiments of the present disclosure do not limit the network type, network structure, and training method of the neural network.

In the embodiment of the present disclosure, the overall similarity between the real human body posture and the recommended human posture can be displayed, so as to effectively guide the user to adjust the posture according to the overall similarity, and improve the shooting experience.

As described above, in step S15, when the photographing conditions are satisfied, a photographed image of the subject is obtained. In a possible implementation manner, the shooting conditions may include that the second similarity between the real human body posture and the recommended human body posture is greater than or equal to a second preset threshold. In this way, the shooting operation can be automatically triggered when the real human body posture is highly similar to the recommended human body posture, and the shooting image can be obtained and saved in a local album for the object to view.

In a possible implementation manner, when the image capturing method is executed by the server, the server may determine that the second similarity between the real human body posture and the recommended human body posture is greater than or equal to a second preset threshold. In this case, a shooting instruction is sent to the terminal device, and the terminal device is triggered to perform a shooting operation through the shooting instruction to obtain a shot image. In the case where the image capturing method is performed by a terminal device, the terminal device may automatically trigger a capturing operation when the second similarity between the real human body posture and the recommended human body posture is greater than or equal to a second preset threshold , to get the captured image.

The second preset threshold may be set according to actual requirements, for example, may be set to 80%, which is not limited in this embodiment of the present disclosure. It should be understood that the shooting conditions may also include that the shooting controls in the shooting interface are triggered, that is, the photographer manually performs the shooting operation.

It is considered that if the object has a better position in the actual scene, or the object is in a suitable position in the scene image, the captured image can present a better visual effect, for example, the composition is more harmonious and pleasing to the eye. In a possible implementation, the method further includes:

Input the scene image to the regional recommendation network to obtain the recommended photographing area of the scene image. The recommended photographing area is used to represent the recommended photographing area in the scene image. The area recommendation network is a neural network obtained by training the first sample set of marking the photographing area. network;

Wherein, after obtaining the recommended photographing area of the scene image, the method further includes at least one of the following:

In the shooting interface, an identifier is used to indicate the recommended shooting area;

The recommended human posture is displayed in the recommended photographing area of the photographing interface, and the recommended photographing area is used to guide the object to adjust the position according to the recommended photographing area.

The region recommendation network may adopt a known neural network, such as a convolutional neural network, etc. The embodiment of the present disclosure does not limit the network type, network structure, and training method of the region recommendation network.

In a possible implementation manner, the sample images of the first sample set may be images with better visual effects. By learning the recommended photographing area from the first sample set with better visual effect, the object can be guided to adjust the position to improve the visual effect of the photographed image, thereby improving the user's photographing experience.

The region where the human body is located in the sample images of the first sample set may be the marked photographing region. It should be understood that, a known labeling technology may be used to realize labeling of the human body region in the sample image of the first sample set, which is not limited in this embodiment of the present disclosure.

In a possible implementation manner, the identifier used to indicate the recommended photographing area may be displayed in the photographing interface in any form such as characters and/or graphics, so as to guide the user object to adjust the position according to the recommended photographing area. The embodiments of the present disclosure are not limited. FIG. 9 shows a schematic diagram of a photographing interface according to an embodiment of the present disclosure. As shown in FIG. 9 , a recommended photographing area may be indicated in the form of graphics, prompt words "recommended station", and the like.

As described above, the recommended human posture may be displayed in a fixed area (the first designated area or the second designated area) of the photographing interface. In a possible implementation manner, the recommended human body posture may be displayed in the recommended photographing area of the photographing interface, in other words, the recommended photographing area may be indicated by the recommended human body posture. It should be understood that the above-mentioned first designated area or second designated area may include a recommended photographing area. In this way, the object can be effectively guided to adjust the position according to the recommended photographing area, and it is also convenient for the object to adjust the posture according to the recommended human body posture.

It should be understood that the recommended photographing area may be indicated by both the identifier and the recommended human posture, or the recommended photographing area may be indicated by only the identifier or the recommended human posture, which is not limited by the embodiment of the present disclosure.

In a possible implementation, when an identifier is used to indicate the recommended photographing area, and it is detected that the object in the scene image is within the recommended photographing area, the identifier can be hidden, thereby making the photographing interface more concise. Wherein, determining whether the object is in the recommended photographing area may be determined according to the position of the object's human body area in the scene image, or may be determined according to the position coordinates of the key points of the human body, which is not limited in this embodiment of the present disclosure.

In the embodiment of the present disclosure, a photographing area with better visual effect can be recommended to the object, and the object can be effectively guided to adjust the position to improve the visual effect of the photographed image, thereby improving the user's photographing experience.

As described above, the scene image can be recognized by the gesture recommendation network, and the scene category corresponding to the scene image and the recommended human pose corresponding to the scene category can be determined. In a possible implementation manner, in step S12, the scene image is identified, and the scene category corresponding to the scene image and the recommended human posture corresponding to the scene category are determined, including:

The scene image is recognized by the gesture recommendation network, and the scene category corresponding to the scene image and the recommended human pose corresponding to the scene category are determined. The gesture recommendation network is obtained by training the second sample set marked with the sample scene category and the sample human pose. The neural network of the second sample set includes sample images uploaded by the object.

Wherein, the pose recommendation network may adopt a known neural network, for example, a convolutional neural network, a residual neural network, and the like. As described above, the embodiments of the present disclosure do not limit the network type, network structure, training method, etc. of the gesture recommendation network.

As mentioned above, the self-supervised training method can be used to make the posture recommendation network learn the recommended human posture under different scene types, or, in other words, learn the correspondence between the scene type and the recommended human posture, so that the posture recommendation network can be used based on Scene image output recommends human pose.

It should be understood that the first sample set and the second sample set may be the same sample set or different sample sets, and the source of the sample images in the first sample set and the second sample set is not limited in the embodiment of the present disclosure. .

Wherein, a known labeling technology can be used to realize labeling of the sample scene category and the sample human body posture in the sample images of the second sample set, which is not limited by the embodiment of the present disclosure.

In a possible implementation manner, an application program for realizing the image capturing method or a capturing mode for realizing the image capturing method may also provide an upload function for uploading a sample image, and the sample image may be an image that has been captured by the user with a relatively low effect. good image. In this way, the second sample set used for training the pose recommendation network can be enriched, so that the trained pose recommendation network can output the recommended human pose based on the scene image more accurately.

It should be understood that the sample image uploaded by the user is the image that the user authorizes to enter the background training posture recommendation network, so the sample image uploaded by the user is only visible to the background, and what other users see are the processed recommended human posture and the corresponding posture. image (for example, the human face is hidden from the gesture image, or the human face is replaced or the character is replaced, etc.), thereby ensuring the privacy and security of the user, wherein the gesture image can be used to select different recommended human gestures.

Wherein, the image uploading function may be implemented in an application program and/or a shooting mode by using a known technology in the art, which is not limited by the embodiment of the present disclosure.

It should be understood that the training process of the gesture recommendation network may be performed on the server. In a possible implementation, the sample images can be uploaded to the server to expand the second sample set, and the posture recommendation network can be periodically retrained based on the second sample set to obtain a new version of the posture recommendation network, and then update the deployed posture recommendation network in the terminal device. The pose recommendation network. In this way, the posture recommendation network can learn rich human body postures, so that the recommended human postures output by the trained posture recommendation network are more comprehensive and richer.

In a possible implementation manner, the retraining process of the posture recommendation network deployed in the terminal device can also be performed on the terminal device, that is, the sample image uploaded by the object can be directly used to retrain the deployed gesture recommendation network on the terminal device. The pose recommendation network is incrementally trained. In this way, the recommended human posture output by the deployed posture recommendation network can be closer to the object's preference and targeted.

It should be understood that the second sample set may be high-quality images with better shooting effects. In the embodiment of the present disclosure, the posture recommendation network can learn the human body posture from the high-quality images, thereby effectively outputting high-quality recommended human postures based on scene images. , to guide the subject to approach the high-quality recommended human posture, so as to obtain a better shooting effect and improve the user's shooting experience.

As described above, there are multiple recommended human body poses. In order to facilitate object selection or switching between different recommended human body poses, in a possible implementation manner, the method further includes:

Displaying a plurality of posture maps corresponding to the recommended human postures in the shooting interface;

In response to the selection operation on the posture map, according to the selected posture map, a recommended human body posture corresponding to the selected posture map is determined and displayed.

FIG. 10 shows a schematic diagram of a photographing interface according to an embodiment of the present disclosure. As shown in FIG. 10 , the photographing interface can display a plurality of posture diagrams, and can also display the recognized scene features “window frame, wall” and objects Features "Sweater, Pants, Hand Raise" for more fun. Wherein, the identified scene features and object features can be obtained by referring to the method of identifying the scene image in the above-mentioned embodiments of the present disclosure. It should be understood that, a sliding control for viewing multiple posture diagrams may also be provided in the photographing interface, so as to view other posture graphs that are not currently displayed in the photographing interface through a sliding operation.

In a possible implementation manner, the displayed posture map may be a thumbnail, which can be understood as a preview map, a schematic diagram, etc. corresponding to the recommended human body posture. It should be understood that the posture map corresponds to the recommended human body posture, and according to any selected posture map, the corresponding recommended human body posture can be obtained.

In a possible implementation manner, the selection operation for the posture map may include, for example, operations such as swiping to view the posture map, clicking to select any posture map, and the like. Among them, the selected posture map can be highlighted to achieve friendly human-computer interaction.

As mentioned above, the displayed posture map can be a thumbnail image. In a possible implementation, after selecting any posture map in the shooting interface as shown in FIG. 10 , the zoomed-in corresponding to the posture map can be displayed in the shooting interface. map, so that the object can view the pose map more clearly.

11a and 11b illustrate schematic diagrams of a photographing interface according to an embodiment of the present disclosure. As shown in Fig. 11a, the enlarged image corresponding to the selected posture image can be displayed in the middle area, wherein, in response to operations such as long-pressing the enlarged image, double-clicking the enlarged image, etc., the image can be previewed in the enlarged image as shown in Fig. 11b. The recommended human pose corresponding to the pose map.

In a possible implementation manner, the selected posture diagram and the corresponding recommended human posture can be determined by clicking the "shoot the same style" button shown in Figure 11a and Figure 11b, and shown in Figure 5, Figure 7, The recommended human posture corresponding to the selected posture map is displayed in the photographing interface shown in FIG. 8 . Wherein, for the manner of displaying the recommended human body posture, reference may be made to the above-mentioned embodiments of the present disclosure.

It should be noted that the manner of displaying the attitude diagram and selecting the attitude diagram in FIG. 10 , FIG. 11 a , and FIG. 11 b above is an implementation method provided by the embodiment of the present disclosure. In fact, those skilled in the art can design the The display manner of the attitude map in the shooting interface, the selection method for the attitude map, etc., are not limited in this embodiment of the present disclosure.

In the embodiment of the present disclosure, it is convenient for the user to select different recommended human body postures, to satisfy different posture preferences, to improve the shooting experience, and to help obtain a photographed image that satisfies the user.

In a possible implementation, the method further includes:

The captured image is processed to obtain a processed captured image, wherein the processing includes at least one of the following: performing beautification processing on an object in the captured image, and adding a filter to the captured image;

The adding a filter to the captured image includes: adjusting at least one of saturation, color temperature, and brightness of the captured image according to a recommended filter corresponding to the recommended human posture.

Wherein, any known image beautifying technology can be used to realize the beautifying processing of the object in the captured image, which is not limited by the embodiment of the present disclosure. For example, the process of beauty processing may include: determining the position of the face in the captured image, then determining the position of the facial defect, and performing processing such as filling, repairing or filtering according to the location of the facial defect.

As shown above, the recommended human pose can be learned from the sample image. In a possible implementation, the recommended filter can be obtained by analyzing the filter of the sample image. Based on this, the recommended human pose can correspond to the recommended filter. .

It should be understood that the recommended filter may include recommended filter parameters such as saturation, color temperature, and brightness. In a possible implementation manner, adjusting at least one of the saturation, color temperature, and brightness of the captured image may include: adjusting at least one of the saturation, color temperature, and brightness of the captured image to match the recommended filter The filter parameters are the same.

Wherein, those skilled in the art can use image processing technologies known in the art to adjust the saturation, color temperature, and brightness of the captured image, which is not limited to the embodiment of the present disclosure.

In the embodiment of the present disclosure, the visual effect of the captured image can be improved, which is beneficial to obtain the captured image that satisfies the user, and improves the image capturing experience.

In a possible implementation manner, the image capturing method of the embodiment of the present disclosure can be applied to artificial intelligence education platforms, social sharing platforms, video image capturing software, etc., and can intelligently recommend recommended human bodies corresponding to the scene images according to the scene images. The posture is used for self-portrait and camera composition guidance for users, which improves the fun of image shooting, improves the shooting experience, and is conducive to obtaining a shooting image that satisfies the user.

Among them, when the image capturing method is applied to the artificial intelligence education platform, intelligent hardware connected to the artificial intelligence education platform can be used, such as Jetson Nano (a microcomputer developed by NVIDIA), Raspberry Pi (a microcomputer) ) etc. to implement the above image capturing method, through this method, based on the form of visualization, it is easier for the user to inspire learning interest.

FIG. 12 shows a schematic diagram of an application of an image capturing method according to an embodiment of the present disclosure. As shown in Figure 12, the artificial intelligence education platform is connected to the intelligent hardware, wherein the artificial intelligence education platform is used to edit the project code to realize the image capturing method, and send the project code to the intelligent hardware, and the intelligent hardware is used to collect scene images And run the project code, get the running results, and send the running results to the artificial intelligence education platform to display the running results in the display interface of the artificial intelligence education platform.

Among them, the scene image, the captured image and the operation result of the image capturing method can be displayed on the display interface of the web page of the artificial intelligence education platform. The running results may include recommended human body postures, real human body postures, captured images after adding filters to the captured images, captured images after beautifying the objects in the captured images, and the like.

In a possible implementation, the artificial intelligence education platform can support students to edit and implement the project code of the above-mentioned image shooting method on the platform, such as training code such as posture recommendation network, and connect with intelligent hardware, which will be used to realize the above-mentioned image shooting. The project code of the method is sent to the intelligent hardware for execution, so that the scene image can be collected in real time by the intelligent hardware, and the above-mentioned image capturing method can be realized based on the collected scene image.

Among them, students can also edit the project code on the artificial intelligence education platform to update and optimize the above-mentioned image shooting methods, such as optimizing each neural network.

In the embodiment of the present disclosure, it is not necessary to install various hardware drivers locally, deploy the dependent libraries and dependent environments of various artificial intelligence algorithms, and enable users to learn through interesting cases, such as human body gesture recognition, human body key point detection, etc. artificial intelligence algorithms.

It can be understood that the above-mentioned method embodiments mentioned in the present disclosure can be combined with each other to form a combined embodiment without violating the principle and logic. Those skilled in the art can understand that, in the above method of the specific embodiment, the specific execution order of each step should be determined by its function and possible internal logic.

In addition, the present disclosure also provides image capturing devices, electronic devices, computer-readable storage media, and programs, all of which can be used to implement any image capturing method provided by the present disclosure. For the corresponding technical solutions and descriptions, refer to the corresponding records in the Methods section. ,No longer.

FIG. 13 shows a block diagram of an image capturing apparatus according to an embodiment of the present disclosure. As shown in FIG. 13 , the apparatus includes:

The acquiring part 101 is configured to acquire the scene image corresponding to the shooting interface;

The identification part 102 is configured to identify the scene image, and determine the scene category corresponding to the scene image and the recommended human posture corresponding to the scene category;

The detection part 103 is configured to perform human body key point detection on the scene image, and determine the object in the scene image and the real human body posture of the object;

The display part 104 is configured to guide the object to perform posture adjustment according to the recommended human body posture by displaying the real human body posture and the recommended human body posture in the shooting interface;

The photographing part 105 is configured to obtain a photographed image of the object when photographing conditions are satisfied.

In a possible implementation manner, the recommended human posture includes a single-person posture, and the display part 104 includes: a first display sub-part, configured to, according to the position of the object in the scene image, display The real human body posture and the single-person posture are displayed in the shooting interface; or, the second display sub-section is configured to display all the objects in the shooting interface according to the position of the object in the scene image. The real human body posture is displayed, and the single-person posture is displayed in the first designated area of the shooting interface.

In a possible implementation manner, the recommended human body posture includes a multi-person combined posture, and the apparatus further includes: a position determination part configured to The relative position between each gesture in the multi-person combined gesture, and the corresponding gesture of each object is respectively determined from the multi-person combined gesture; wherein, the display part 104 includes: a third display sub-section, which is configured for For any object, according to the position of the object in the scene image, the real human body posture of the object and the corresponding posture of the object are displayed in the shooting interface; or, the fourth display subsection is configured as According to the position of the object in the scene image, the real human body posture of the object is displayed in the shooting interface, and the corresponding posture of the object is displayed in the second designated area of the shooting interface.

In a possible implementation manner, the apparatus further includes: a first determining part configured to determine the distance between the first human body key point of the real human body posture and the second human body key point of the recommended human body posture The key point pair; the second determination part is configured to determine from the first human body key points that the first similarity is less than the first preset threshold by calculating the first similarity of the key point pair. The third human body key point; wherein, the display part 104 includes: a highlighting sub-section configured to highlight the region where the third human body key point is located, wherein the highlighting method includes highlighting, bolding, Change at least one of the colors.

In a possible implementation manner, the apparatus further includes: a third determination part, configured to determine the the second similarity between the real human body posture and the recommended human body posture, and the key point pair is determined according to the first human body key point of the real human body posture and the second human body key point of the recommended human body posture; The similarity display part is configured to display the second similarity in the shooting interface.

In a possible implementation manner, the apparatus further includes: an area determination part, configured to input the scene image into an area recommendation network to obtain a recommended photographing area of the scene image, where the recommended photographing area is used for Representing the area recommended to be photographed in the scene image, and the area recommendation network is a neural network trained by marking the first sample set of the photographing area; wherein, the device further includes: an area display part, configured to In the shooting interface, an identifier is used to indicate the recommended photographing area; or, the recommended human posture is displayed in the recommended photographing area of the photographing interface; the recommended photographing area is used to guide the object according to the recommended photographing area. Adjust the position of the recommended photo area as described above.

In a possible implementation manner, the area display part is further configured to use an identifier to indicate the recommended photographing area in the photographing interface; and display the recommended photographing area on the photographing interface. The recommended human posture is described.

In a possible implementation manner, the identifying part 102 is further configured to identify the scene image through a gesture recommendation network, and determine a scene category corresponding to the scene image and a recommended human body corresponding to the scene category pose, wherein the pose recommendation network is a neural network obtained by training a second sample set marked with sample scene categories and sample human poses, and the second sample set includes sample images uploaded by objects.

In a possible implementation manner, the recommended human body postures include a plurality of postures, and the apparatus further includes: a posture map display part configured to display the posture maps corresponding to the plurality of recommended human body postures in the shooting interface; selecting The part is configured to, in response to a selection operation on the posture map, determine and display a recommended human body posture corresponding to the selected posture map according to the selected posture map.

In a possible implementation manner, the shooting conditions include: a second degree of similarity between the real human body posture and the recommended human body posture is greater than or equal to a second preset threshold; or, in the shooting interface Shooting controls are triggered.

In a possible implementation manner, the apparatus further includes: a processing part configured to process the captured image to obtain a processed captured image, wherein the processing includes at least one of the following: Performing beautification processing on the object in the captured image, and adding a filter to the captured image; the adding a filter to the captured image includes: adjusting the captured image according to the recommended filter corresponding to the recommended human posture at least one of saturation, color temperature, and brightness.

In a possible implementation, the device is applied to an artificial intelligence education platform, and the artificial intelligence education platform is connected to intelligent hardware; wherein, the artificial intelligence education platform is used to edit and implement the project code of the image capturing device , and send the project code to the intelligent hardware; the intelligent hardware is used to collect scene images and run the project code to obtain the operation result, and send the operation result to the artificial intelligence education platform to The operation result is displayed on the display interface of the artificial intelligence education platform.

In the embodiment of the present disclosure, the recommended human body pose corresponding to the scene category can be recommended to the user according to the scene category of the scene image, so that the recommended human body pose matches the scene image, and by displaying the real human body pose and The recommended human posture can guide the user to adjust the posture according to the displayed recommended human posture or facilitate the user to guide the subject to adjust the posture according to the displayed recommended human posture. Satisfied taking pictures.

In some embodiments, the functions or included parts of the apparatus provided in the embodiments of the present disclosure may be configured to execute the methods described in the above method embodiments, and the specific implementation may refer to the above method embodiments. For brevity, I won't go into details here.

Embodiments of the present disclosure further provide a computer-readable storage medium, on which computer program instructions are stored, and when the computer program instructions are executed by a processor, the foregoing method is implemented. Computer-readable storage media can be volatile or non-volatile computer-readable storage media.

An embodiment of the present disclosure further provides an electronic device, including: a processor; a memory configured to store instructions executable by the processor; wherein the processor is configured to invoke the instructions stored in the memory to execute the above method.

Embodiments of the present disclosure also provide a computer program product, including computer-readable codes. When the computer-readable codes are run on a device, a processor in the device executes the image capturing method for implementing the image capturing method provided by any of the above embodiments. instruction.

Embodiments of the present disclosure further provide another computer program product for storing computer-readable instructions, which, when executed, cause the computer to perform the operations of the image capturing method provided by any of the foregoing embodiments.

The electronic device in the embodiment of the present disclosure may be provided as a terminal device or a server.

FIG. 14 shows a block diagram of an electronic device 800 according to an embodiment of the present disclosure. For example, electronic device 800 may be a mobile phone, computer, digital broadcast terminal, messaging device, game console, tablet device, medical device, fitness device, personal digital assistant, etc. terminal.

14, an electronic device 800 may include one or more of the following components: a processing component 802, a memory 804, a power supply component 806, a multimedia component 808, an audio component 810, an input/output (I/O) interface 812, a sensor component 814 , and the communication component 816 .

The processing component 802 generally controls the overall operation of the electronic device 800, such as operations associated with display, phone calls, data communications, camera operations, and recording operations. The processing component 802 can include one or more processors 820 to execute instructions to perform all or some of the steps of the methods described above. Additionally, processing component 802 may include one or more modules that facilitate interaction between processing component 802 and other components. For example, processing component 802 may include a multimedia module to facilitate interaction between multimedia component 808 and processing component 802.

Memory 804 is configured to store various types of data to support operation at electronic device 800 . Examples of such data include instructions for any application or method operating on electronic device 800, contact data, phonebook data, messages, pictures, videos, and the like. Memory 804 may be implemented by any type of volatile or nonvolatile storage device or combination thereof, such as static random access memory (SRAM), electrically erasable programmable read only memory (EEPROM), erasable Programmable Read Only Memory (EPROM), Programmable Read Only Memory (PROM), Read Only Memory (ROM), Magnetic Memory, Flash Memory, Magnetic or Optical Disk.

Power supply assembly 806 provides power to various components of electronic device 800 . Power supply components 806 may include a power management system, one or more power supplies, and other components associated with generating, managing, and distributing power to electronic device 800 .

Multimedia component 808 includes a screen that provides an output interface between the electronic device 800 and the user. In some embodiments, the screen may include a liquid crystal display (LCD) and a touch panel (TP). If the screen includes a touch panel, the screen may be implemented as a touch screen to receive input signals from a user. The touch panel includes one or more touch sensors to sense touch, swipe, and gestures on the touch panel. The touch sensor may not only sense the boundaries of a touch or swipe action, but also detect the duration and pressure associated with the touch or swipe action. In some embodiments, the multimedia component 808 includes a front-facing camera and/or a rear-facing camera. When the electronic device 800 is in an operation mode, such as a shooting mode or a video mode, the front camera and/or the rear camera may receive external multimedia data. Each of the front and rear cameras can be a fixed optical lens system or have focal length and optical zoom capability.

Audio component 810 is configured to output and/or input audio signals. For example, audio component 810 includes a microphone (MIC) that is configured to receive external audio signals when electronic device 800 is in operating modes, such as calling mode, recording mode, and voice recognition mode. The received audio signal may be further stored in memory 804 or transmitted via communication component 816 . In some embodiments, audio component 810 also includes a speaker for outputting audio signals.

The I/O interface 812 provides an interface between the processing component 802 and a peripheral interface module, which may be a keyboard, a click wheel, a button, or the like. These buttons may include, but are not limited to: home button, volume buttons, start button, and lock button.

Sensor assembly 814 includes one or more sensors for providing status assessment of various aspects of electronic device 800 . For example, the sensor assembly 814 can detect the on/off state of the electronic device 800, the relative positioning of the components, such as the display and the keypad of the electronic device 800, the sensor assembly 814 can also detect the electronic device 800 or one of the electronic device 800 Changes in the position of components, presence or absence of user contact with the electronic device 800 , orientation or acceleration/deceleration of the electronic device 800 and changes in the temperature of the electronic device 800 . Sensor assembly 814 may include a proximity sensor configured to detect the presence of nearby objects in the absence of any physical contact. Sensor assembly 814 may also include a light sensor, such as a complementary metal oxide semiconductor (CMOS) or charge coupled device (CCD) image sensor, for use in imaging applications. In some embodiments, the sensor assembly 814 may also include an acceleration sensor, a gyroscope sensor, a magnetic sensor, a pressure sensor, or a temperature sensor.

Communication component 816 is configured to facilitate wired or wireless communication between electronic device 800 and other devices. The electronic device 800 may access a wireless network based on a communication standard, such as wireless network (WiFi), second generation mobile communication technology (2G) or third generation mobile communication technology (3G), or a combination thereof. In one exemplary embodiment, the communication component 816 receives broadcast signals or broadcast related information from an external broadcast management system via a broadcast channel. In an exemplary embodiment, the communication component 816 also includes a near field communication (NFC) module to facilitate short-range communication. For example, the NFC module may be implemented based on radio frequency identification (RFID) technology, infrared data association (IrDA) technology, ultra-wideband (UWB) technology, Bluetooth (BT) technology and other technologies.

In an exemplary embodiment, electronic device 800 may be implemented by one or more application specific integrated circuits (ASICs), digital signal processors (DSPs), digital signal processing devices (DSPDs), programmable logic devices (PLDs), field programmable A programmed gate array (FPGA), controller, microcontroller, microprocessor or other electronic component implementation is used to perform the above method.

In an exemplary embodiment, a non-volatile computer-readable storage medium, such as a memory 804 comprising computer program instructions executable by the processor 820 of the electronic device 800 to perform the above method is also provided.

The present disclosure may be a system, method and/or computer program product. The computer program product may include a computer-readable storage medium having computer-readable program instructions loaded thereon for causing a processor to implement various aspects of the present disclosure.

A computer-readable storage medium may be a tangible device that can hold and store instructions for use by the instruction execution device. The computer-readable storage medium may be, for example, but not limited to, an electrical storage device, a magnetic storage device, an optical storage device, an electromagnetic storage device, a semiconductor storage device, or any suitable combination of the foregoing. More specific examples (non-exhaustive list) of computer readable storage media include: portable computer disks, hard disks, random access memory (RAM), read only memory (ROM), erasable programmable read only memory (EPROM) or flash memory), static random access memory (SRAM), portable compact disk read only memory (CD-ROM), digital versatile disk (DVD), memory sticks, floppy disks, mechanically coded devices, such as printers with instructions stored thereon Hole cards or raised structures in grooves, and any suitable combination of the above. Computer-readable storage media, as used herein, are not to be construed as transient signals per se, such as radio waves or other freely propagating electromagnetic waves, electromagnetic waves propagating through waveguides or other transmission media (eg, light pulses through fiber optic cables), or through electrical wires transmitted electrical signals.

The computer readable program instructions described herein may be downloaded to various computing/processing devices from a computer readable storage medium, or to an external computer or external storage device over a network such as the Internet, a local area network, a wide area network, and/or a wireless network. The network may include copper transmission cables, fiber optic transmission, wireless transmission, routers, firewalls, switches, gateway computers, and/or edge servers. A network adapter card or network interface in each computing/processing device receives computer-readable program instructions from a network and forwards the computer-readable program instructions for storage in a computer-readable storage medium in each computing/processing device .

Computer program instructions for carrying out operations of the present disclosure may be assembly instructions, instruction set architecture (ISA) instructions, machine instructions, machine-dependent instructions, microcode, firmware instructions, state setting data, or instructions in one or more programming languages. Source or object code, written in any combination, including object-oriented programming languages, such as Smalltalk, C++, etc., and conventional procedural programming languages, such as the "C" language or similar programming languages. The computer readable program instructions may execute entirely on the user's computer, partly on the user's computer, as a stand-alone software package, partly on the user's computer and partly on a remote computer, or entirely on the remote computer or server implement. In the case of a remote computer, the remote computer may be connected to the user's computer through any kind of network, including a local area network (LAN) or a wide area network (WAN), or may be connected to an external computer (eg, using an Internet service provider through the Internet connect). In some embodiments, custom electronic circuits, such as programmable logic circuits, field programmable gate arrays (FPGAs), or programmable logic arrays (PLAs), can be personalized by utilizing state information of computer readable program instructions. Computer readable program instructions are executed to implement various aspects of the present disclosure.

Aspects of the present disclosure are described herein with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems) and computer program products according to embodiments of the disclosure. It will be understood that each block of the flowchart illustrations and/or block diagrams, and combinations of blocks in the flowchart illustrations and/or block diagrams, can be implemented by computer readable program instructions.

These computer readable program instructions may be provided to a processor of a general purpose computer, special purpose computer or other programmable data processing apparatus to produce a machine that causes the instructions when executed by the processor of the computer or other programmable data processing apparatus , resulting in means for implementing the functions/acts specified in one or more blocks of the flowchart and/or block diagrams. These computer readable program instructions can also be stored in a computer readable storage medium, these instructions cause a computer, programmable data processing apparatus and/or other equipment to operate in a specific manner, so that the computer readable medium storing the instructions includes An article of manufacture comprising instructions for implementing various aspects of the functions/acts specified in one or more blocks of the flowchart and/or block diagrams.

Computer readable program instructions can also be loaded onto a computer, other programmable data processing apparatus, or other equipment to cause a series of operational steps to be performed on the computer, other programmable data processing apparatus, or other equipment to produce a computer-implemented process , thereby causing instructions executing on a computer, other programmable data processing apparatus, or other device to implement the functions/acts specified in one or more blocks of the flowcharts and/or block diagrams.

The flowchart and block diagrams in the Figures illustrate the architecture, functionality, and operation of possible implementations of systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each block in the flowchart or block diagrams may represent a module, segment, or portion of instructions, which comprises one or more functions for implementing the specified logical function(s) executable instructions. In some alternative implementations, the functions noted in the blocks may occur out of the order noted in the figures. For example, two blocks in succession may, in fact, be executed substantially concurrently, or the blocks may sometimes be executed in the reverse order, depending upon the functionality involved. It is also noted that each block of the block diagrams and/or flowchart illustrations, and combinations of blocks in the block diagrams and/or flowchart illustrations, can be implemented in dedicated hardware-based systems that perform the specified functions or actions , or can be implemented in a combination of dedicated hardware and computer instructions.

The computer program product can be specifically implemented by hardware, software or a combination thereof. In an optional embodiment, the computer program product is embodied as a computer storage medium, and in another optional embodiment, the computer program product is embodied as a software product, such as a software development kit (Software Development Kit, SDK), etc. Wait.

Various embodiments of the present disclosure have been described above, and the foregoing descriptions are exemplary, not exhaustive, and not limiting of the disclosed embodiments. Numerous modifications and variations will be apparent to those of ordinary skill in the art without departing from the scope and spirit of the described embodiments. The terminology used herein was chosen to best explain the principles of the various embodiments, the practical application or improvement over the technology in the marketplace, or to enable others of ordinary skill in the art to understand the various embodiments disclosed herein.

Industrial Applicability

In the embodiment of the present disclosure, the recommended human body pose corresponding to the scene category can be recommended to the user according to the scene category of the scene image, so that the recommended human body pose matches the scene image, and by displaying the real human body pose and The recommended human posture can guide the user to adjust the posture according to the displayed recommended human posture or facilitate the user to guide the subject to adjust the posture according to the displayed recommended human posture. Satisfied taking pictures. And, by displaying the single-person posture on the shooting interface, the subject can be effectively guided to complete the shooting of the single-person posture, improving the shooting experience, and it is recommended that the human body posture can be displayed fixedly and can be displayed with the object, which can meet different posture display needs. And, by displaying the combined pose of multiple people on the shooting interface, multiple objects can be effectively guided to complete the combined pose of multiple people, improving the shooting experience, and it is recommended that the human body pose can be displayed fixedly and can be displayed with the object, which can meet different pose display needs. And, the parts of the human body with low similarity among the displayed real human body postures can be highlighted, so as to more effectively guide the user to adjust the posture, so as to improve the shooting experience. And, the overall similarity between the real human body posture and the recommended human posture can be displayed, so as to effectively guide the user to adjust the posture according to the overall similarity, and improve the shooting experience. And, by recommending a photographing area, a photographing area with better visual effect can be recommended to the subject, and the subject can be effectively guided to adjust the position to improve the visual effect of the photographed image, thereby improving the user's photographing experience. And, the pose recommendation network can be used to learn human poses from high-quality images, so as to effectively output high-quality recommended human poses based on scene images. And it is convenient for the user to select different recommended human body postures, to satisfy different posture preferences, to improve the shooting experience, and to be beneficial to obtain a shooting image that satisfies the user. And, applying the methods in the embodiments of the present disclosure in an artificial intelligence education platform enables users to learn through interesting cases, and learn artificial intelligence algorithms such as human body gesture recognition and human body key point detection.

Claims (25)

An image capturing method comprising: Obtain the scene image corresponding to the shooting interface; Identifying the scene image, and determining a scene category corresponding to the scene image and a recommended human posture corresponding to the scene category; Performing human body key point detection on the scene image, and determining the object in the scene image and the real human body posture of the object; By displaying the real human body posture and the recommended human body posture in the shooting interface, the object is guided to perform posture adjustment according to the recommended human body posture; When the shooting conditions are satisfied, a shot image of the object is obtained. The method according to claim 1, wherein the recommended human body posture includes a single person posture, and the displaying the real human body posture and the recommended human body posture in the shooting interface comprises: According to the position of the object in the scene image, the real human body posture and the single-person posture are displayed in the shooting interface; or, According to the position of the object in the scene image, the real human pose is displayed in the shooting interface, and the single-person pose is displayed in a first designated area of the shooting interface. The method according to claim 1 or 2, wherein the recommended human body posture comprises a multi-person combined posture, and the method further comprises: According to the relative positions between the multiple objects in the scene image and the relative positions between the respective gestures in the multiple-person combined gestures, the corresponding gestures of each object are respectively determined from the multiple-person combined gestures; Wherein, displaying the real human body posture and the recommended human body posture in the shooting interface includes: For any object, display the real human body posture of the object and the corresponding posture of the object in the shooting interface according to the position of the object in the scene image; or, According to the position of the object in the scene image, the real human body posture of the object is displayed in the shooting interface, and the corresponding posture of the object is displayed in the second designated area of the shooting interface. The method according to any one of claims 1-3, wherein the method further comprises: Determine the key point pair between the first human body key point of the real human body posture and the second human body key point of the recommended human body posture; By calculating the first similarity of the key point pair, a third human body key point whose first similarity is less than a first preset threshold is determined from the first human body key points; Wherein, the displaying the real human body posture and the recommended human body posture in the shooting interface includes: The region where the third human body key point is located is highlighted, and the highlighting manner includes at least one of highlighting, bolding, and changing color. The method according to any one of claims 1-4, wherein the method further comprises: According to the first similarity of the key point pair between the real human body posture and the recommended human body posture, the second similarity between the real human body posture and the recommended human body posture is determined, and the key point pair is Determined according to the first human body key point of the real human body posture and the second human body key point of the recommended human body posture; The second similarity is displayed in the shooting interface. The method according to any one of claims 1-5, wherein the method further comprises: Inputting the scene image into a regional recommendation network to obtain a recommended photographing area of the scene image, where the recommended photographing area is used to represent the recommended photographing area in the scene image, and the area recommending network is to mark the photographing area by marking the photographing area. The neural network obtained by training the first sample set of ; Wherein, after obtaining the recommended photographing area of the scene image, the method further includes at least one of the following: In the photographing interface, an identifier is used to indicate the recommended photographing area; The recommended human posture is displayed in the recommended photographing area of the photographing interface; the recommended photographing area is used to guide the object to perform position adjustment according to the recommended photographing area. The method according to any one of claims 1-6, wherein identifying the scene image, determining a scene category corresponding to the scene image and a recommended human pose corresponding to the scene category, comprising: Identify the scene image through a gesture recommendation network, and determine the scene category corresponding to the scene image and the recommended human pose corresponding to the scene category, wherein the gesture recommendation network is marked with the sample scene category and the sample human body. A neural network obtained by training a second sample set of poses, where the second sample set includes sample images uploaded by the object. The method according to claim 1, wherein the recommended human body postures include multiple ones, and the method further comprises: Displaying a plurality of posture diagrams corresponding to the recommended human postures in the shooting interface; In response to the selection operation on the posture map, according to the selected posture map, a recommended human body posture corresponding to the selected posture map is determined and displayed. The method according to claim 1 or 5, wherein the shooting conditions include: The second similarity between the real human body posture and the recommended human body posture is greater than or equal to a second preset threshold; or, The shooting controls in the shooting interface are triggered. The method according to any one of claims 1-9, wherein the method further comprises: The captured image is processed to obtain a processed captured image, wherein the processing includes at least one of the following: performing beautification processing on an object in the captured image, and adding a filter to the captured image; The adding a filter to the captured image includes: adjusting at least one of saturation, color temperature, and brightness of the captured image according to a recommended filter corresponding to the recommended human posture. The method according to any one of claims 1-10, wherein the method is applied to an artificial intelligence education platform, and the artificial intelligence education platform is connected with intelligent hardware; Wherein, the artificial intelligence education platform is used to edit the project code for realizing the image capturing method, and send the project code to the intelligent hardware; The intelligent hardware is used to collect scene images and run the project code to obtain the running results, and send the running results to the artificial intelligence education platform to display the artificial intelligence education platform on the display interface. operation result. An image capturing device, comprising: an acquisition part, configured to acquire a scene image corresponding to the shooting interface; an identification part, configured to identify the scene image, and determine a scene category corresponding to the scene image and a recommended human posture corresponding to the scene category; a detection part, configured to perform human body key point detection on the scene image, and determine the object in the scene image and the real human body posture of the object; a display part, configured to display the real human body posture and the recommended human body posture in the shooting interface, so as to guide the object to perform posture adjustment according to the recommended human body posture; The photographing part is configured to obtain a photographed image of the object when photographing conditions are satisfied. The apparatus according to claim 12, wherein the recommended human posture comprises a single-person posture, and the display part comprises: a first display subsection, configured to display the real human body posture and the single-person posture in the shooting interface according to the position of the object in the scene image; or, The second display subsection is configured to display the real human body posture in the shooting interface according to the position of the object in the scene image, and display the real human body posture in the first designated area of the shooting interface Solo pose. The device according to claim 12 or 13, wherein the recommended human body posture comprises a multi-person combined posture, and the device further comprises: a position determination part configured to respectively determine each object from the multi-person combined gesture according to the relative positions between the plurality of objects in the scene image and the relative positions between the various gestures in the multi-person combined gesture The corresponding posture of ; where, The display part includes: a third display sub-part, which is configured to, for any object, display the real human body posture of the object and all the objects in the shooting interface according to the position of the object in the scene image. the corresponding pose of the object; or, a fourth display subsection, configured to display the real human body posture of the object in the shooting interface according to the position of the object in the scene image, and display it at the second designated area of the shooting interface the corresponding pose of the object. The apparatus of any one of claims 12-14, wherein the apparatus further comprises: a first determining part, configured to determine a key point pair between the first human body key point of the real human body posture and the second human body key point of the recommended human body posture; The second determination part is configured to determine, from the first human body key points, a third human body key point whose first similarity is less than a first preset threshold by calculating the first similarity of the key point pair ; Wherein, the display part includes: The highlighting subsection is configured to highlight the region where the third human body key point is located, wherein the highlighting manner includes at least one of highlighting, bolding, and changing color. The apparatus of any one of claims 12-15, wherein the apparatus further comprises: A third determining part is configured to determine a second similarity between the real human body posture and the recommended human body posture according to the first similarity of key point pairs between the real human body posture and the recommended human body posture The key point pair is determined according to the first human body key point of the real human body posture and the second human body key point of the recommended human body posture; The similarity display part is configured to display the second similarity in the shooting interface. The apparatus of any one of claims 12-16, wherein the apparatus further comprises: an area determination part, configured to input the scene image into an area recommendation network to obtain a recommended photographing area of the scene image, where the recommended photographing area is used to represent a recommended photographing area in the scene image, the area The recommendation network is a neural network trained by marking the first sample set of the photo area; Wherein, the device also includes: an area display part, configured to use an identifier to indicate the recommended photographing area in the photographing interface; Or, displaying the recommended human posture in the recommended photographing area of the photographing interface; the recommended photographing area is used to guide the object to perform position adjustment according to the recommended photographing area; Alternatively, the area display part is further configured to use an identifier to indicate the recommended photographing area in the photographing interface; and display the recommended human posture at the recommended photographing area on the photographing interface. The apparatus of any of claims 12-17, wherein, The identification part is further configured to: identify the scene image through a gesture recommendation network, and determine a scene category corresponding to the scene image and a recommended human posture corresponding to the scene category, wherein the gesture recommendation network It is a neural network obtained by training a second sample set marked with sample scene categories and sample body poses, and the second sample set includes sample images uploaded by objects. The device according to claim 12, wherein the recommended human body posture comprises a plurality of, and the device further comprises: The posture map display part is configured to display a plurality of posture maps corresponding to the recommended human body postures in the shooting interface; the selection part is configured to respond to the selection operation for the posture map, according to the selected posture map, determine and The recommended human body posture corresponding to the selected posture map is displayed and displayed. The apparatus according to claim 12 or 16, wherein the shooting condition comprises: a second similarity between the real human body posture and the recommended human body posture is greater than or equal to a second preset threshold; or, the The capture controls in the capture interface are triggered. The apparatus of any one of claims 12-20, wherein the apparatus further comprises: The processing part is configured to process the captured image to obtain a processed captured image, wherein the processing includes at least one of the following: performing beauty processing on the object in the captured image, performing processing on the captured image Adding a filter; the adding a filter to the captured image includes: adjusting at least one of saturation, color temperature, and brightness of the captured image according to a recommended filter corresponding to the recommended human posture. The device according to any one of claims 12-21, wherein the device is applied to an artificial intelligence education platform, and the artificial intelligence education platform is connected with intelligent hardware; wherein, the artificial intelligence education platform is used to edit and implement all the project code of the image capturing device, and send the project code to the intelligent hardware; the intelligent hardware is used to collect scene images and run the project code to obtain the operation result, and send the operation result to the intelligent hardware. The artificial intelligence education platform is used to display the operation result in the display interface of the artificial intelligence education platform. An electronic device comprising: processor; a memory configured to store processor-executable instructions; wherein the processor is configured to invoke instructions stored in the memory to perform the method of any one of claims 1-11. A computer-readable storage medium having computer program instructions stored thereon, the computer program instructions implementing the method of any one of claims 1 to 11 when executed by a processor. A computer program, comprising computer-readable codes, when the computer-readable codes are executed in an electronic device, a processor in the electronic device implements the method described in any one of claims 1 to 11 when executed. method.

Images