CN114693629A - Image recognition method and device, electronic equipment and readable medium - Google Patents

Abstract

The embodiment of the invention provides an image identification method, an image identification device, electronic equipment and a readable medium, wherein the method comprises the following steps: acquiring a target image submitted to an application program; extracting contour edge lines from the target image; determining an edge region in the target image based on the contour edge line; determining the color distribution condition of the edge area; when the color distribution condition meets a preset color distribution condition, identifying the target image as a screen shot image; wherein the color distribution condition is based on a background correspondence setting when an image is viewed in the application program. According to the embodiment of the invention, whether the target image is the screen capture image is determined by automatically identifying whether the edge area of the target image meets the color distribution condition, so that the identification efficiency of the screen capture image is improved.

Description

Image recognition method, device, electronic device and readable medium

technical field

Embodiments of the present invention relate to the field of Internet technologies, and in particular, to an image recognition method, an image recognition apparatus, an electronic device, and a computer-readable medium.

Background technique

With the development of Internet technology, users can realize online shopping, house transaction, job search and other services through applications (Application, APP) installed in terminal devices such as smartphones and computers, which not only greatly improves the convenience of people's life, but also enrich the diversity of life.

When the user uses the various services provided by the application, the image can be used to determine whether to conduct online shopping or house transaction, etc. However, the number of posts in the application is huge, and there may be posts that include images that embezzle legitimate users. One of the common ways of stealing images is to take screenshots, and the images obtained by taking screenshots are screenshots.

Currently, manual review is used to identify whether an image uploaded by a user is a screenshot image. However, relying solely on manual review has the disadvantage of high cost. In addition, due to the low efficiency of manual recognition, facing the huge number of posts in the application, it is impossible to review all images uploaded by users in a timely manner. , so there is a risk of a lot of screenshots appearing online, which in turn affects the user experience.

SUMMARY OF THE INVENTION

Embodiments of the present invention provide an image recognition method, apparatus, electronic device, and computer-readable storage medium, so as to solve the problem of inability to review screenshot images in time.

The embodiment of the present invention discloses an image recognition method, and the method includes:

Get the target image submitted to the application;

extracting contour edge lines from the target image;

determining edge regions in the target image based on the contour edge lines;

determining the color distribution of the edge region;

When the color distribution condition satisfies a preset color distribution condition, the target image is identified as a screenshot image; wherein, the color distribution condition is based on the background corresponding setting when viewing the image in the application program.

Optionally, the determining an edge area in the target image based on the contour edge line includes:

respectively acquiring at least two pixel coordinates of the contour edge line and the distance between the contour edge line and the edge of the target image;

Determine the angle of the contour edge line according to the pixel coordinates;

Determining a preset number of the contour edge lines whose angle is a preset angle and the distance from the edge of the target image is the smallest as the target contour edge line;

The target image between the target contour edge line and the edge of the target image is determined as an edge area.

Optionally, after determining the angle of the contour edge line according to the pixel coordinates, the method further includes:

When there is no contour edge line whose angle is a preset angle, the target image is identified as a non-screenshot image.

Optionally, the determining the color distribution of the edge region includes:

obtaining the pixel value of the pixel in the edge area;

The pixel value is used to calculate the mean and variance of the edge area; the mean and the variance are used to characterize the color distribution of the edge area.

Optionally, when the color distribution condition satisfies a preset color distribution condition, identifying the target image as a screenshot image includes:

When the mean value and the variance satisfy a preset color distribution condition corresponding to a specified color background, and the target image is not a floor plan, the target image is identified as a screenshot image.

Optionally, determine whether the target image is a floor plan in the following manner:

Obtain the image information of the target image submitted in the application;

Whether the target image is a floor plan is determined based on the image information.

Optionally, after the determining the color distribution of the edge region, the method further includes:

When the color distribution condition does not meet the preset color distribution condition, the target image is identified as a non-screenshot image.

Optionally, after the target image is identified as a screenshot image when the color distribution condition satisfies a preset color distribution condition, the method further includes:

If the target image is identified as a screenshot image, publishing a post containing the target image in the application is prohibited.

The embodiment of the present invention also discloses an image recognition device, the device includes:

The target image acquisition module is used to acquire the target image submitted to the application;

an outline edge line extraction module for extracting outline edge lines from the target image;

an edge area determination module, configured to determine an edge area in the target image based on the contour edge line;

a color distribution determination module, used to determine the color distribution of the edge area;

An image recognition module for recognizing the target image as a screenshot image when the color distribution condition satisfies a preset color distribution condition; wherein, the color distribution condition is based on the background when viewing the image in the application program corresponding settings.

Optionally, the edge area determination module is configured to separately obtain at least two pixel coordinates of the contour edge line and the distance between the contour edge line and the edge of the target image; according to the pixel coordinates determining the angle of the contour edge line; determining the preset number of contour edge lines whose angle is a preset angle and the distance from the edge of the target image is the smallest as the target contour edge line; The target image between the target contour edge line and the edge of the target image is determined as an edge area.

Optionally, the edge area determination module is configured to identify the target image as a non-screenshot image when there is no contour edge line whose angle is a preset angle.

Optionally, the color distribution determination module is used to obtain the pixel value of the pixel in the edge area; the pixel value is used to calculate the mean value and variance of the edge area; the mean value and the variance are used to represent The color distribution of the edge area.

Optionally, the image recognition module is configured to classify the target image when the mean value and the variance satisfy preset color distribution conditions set corresponding to the background of the specified color, and the target image is not a floor plan. Recognized as a screenshot image.

Optionally, the image recognition module is configured to acquire image information of the target image submitted in the application; determine whether the target image is a floor plan based on the image information.

Optionally, the image recognition module is configured to recognize the target image as a non-screenshot image when the color distribution does not meet a preset color distribution condition.

Optionally, the apparatus further includes a publishing prohibiting module, configured to prohibit publishing a post containing the target image in the application program if the target image is identified as a screenshot image.

An embodiment of the present invention further discloses an electronic device, including a processor, a communication interface, a memory, and a communication bus, wherein the processor, the communication interface, and the memory communicate with each other through the communication bus;

the memory for storing computer programs;

The processor is configured to implement the method according to the embodiment of the present invention when executing the program stored in the memory.

The embodiments of the present invention also disclose one or more computer-readable media, on which instructions are stored, and when executed by one or more processors, cause the processors to perform the method according to the embodiments of the present invention.

The embodiment of the present invention also discloses a computer program product, the program product is stored in a storage medium, and the computer program product is executed by at least one processor to implement the method according to the embodiment of the present invention.

The embodiments of the present invention include the following advantages:

In the embodiment of the present invention, the target image submitted to the application is acquired, the contour edge line is extracted from the target image, and after determining the edge area in the target image based on the contour edge line, the color distribution of the edge area is determined, if the color distribution If the condition satisfies the preset color distribution condition, the target image can be identified as a screenshot image. The color distribution condition is based on the corresponding setting of the background when viewing the image in the application. If the color distribution of the edge area satisfies the preset color distribution condition , indicating that the target image may be a screenshot image with a screenshot of the background. The embodiment of the present invention determines whether the target image is a screenshot image by automatically identifying whether the edge area of the target image satisfies the color distribution condition, thereby improving the recognition efficiency of the screenshot image.

Description of drawings

1 is a schematic diagram of an application environment provided in an embodiment of the present invention;

2 is a flow chart of steps of an image recognition method provided in an embodiment of the present invention;

3 is one of the screenshot images of a house image provided in an embodiment of the present invention;

4 is the second screenshot of a house image provided in an embodiment of the present invention;

FIG. 5 is the third screenshot image of a house image provided in an embodiment of the present invention;

6 is a schematic diagram of a contour edge line of a house image provided in an embodiment of the present invention;

7 is a schematic diagram of a house type diagram provided in an embodiment of the present invention;

8 is a flowchart of an image recognition provided in an embodiment of the present invention;

FIG. 9 is a structural block diagram of an image recognition apparatus provided in an embodiment of the present invention.

Detailed ways

In order to make the above objects, features and advantages of the present invention more clearly understood, the present invention will be described in further detail below with reference to the accompanying drawings and specific embodiments.

The image recognition method provided by the embodiment of the present invention can be applied to the application environment shown in FIG. 1 . The terminal devices 102 all communicate with the server 104 through the network. Specifically, the server 104 obtains the target image submitted by the user to the application program through the terminal device 102, extracts the contour edge line from the target image, determines the edge area in the target image based on the contour edge line, determines the color distribution of the edge area, and When the color distribution condition satisfies the preset color distribution condition, the target image is identified as a screenshot image; wherein, the color distribution condition is based on the corresponding setting of the background when viewing the image in the application program.

In practical applications, the terminal device 102 can be, but is not limited to, various personal computers, notebook computers, smart phones, tablet computers, robots, vehicle-mounted terminals and portable wearable devices, and the server 104 can be an independent server or composed of multiple servers. server cluster to achieve.

Referring to FIG. 2, a flowchart of steps of an image recognition method provided in an embodiment of the present invention is shown, which may specifically include the following steps:

Step 202: Acquire the target image submitted to the application.

In a specific implementation, an application program, such as a life application program, can be run on the terminal device. Through life applications, users can be provided with house transaction functions, vehicle purchase and sale functions, housekeeping service functions and leisure functions.

Wherein, the target image may be an image containing a specified object. For example, a house image including a house, a vehicle image including a vehicle, an item image including an item, and the like.

For example, a landlord user or a broker user can upload a post about a house transaction in a life application. In order to make the tenant user who browses the post better understand the situation of the house, the post usually contains house images, such as about the house. A photo of the environment, if the post is allowed to be published on the app, then the guest user can browse to the post to determine whether to further communicate with the user who posted the post. In this embodiment of the present invention, if it is detected that a user uploads a house transaction post in a life application program, and the house transaction post contains a house image, the house image in the post can be used as the target image.

Step 204, extracting contour edge lines from the target image.

Among them, the contour edge line, also known as "external line", refers to the outer edge boundary of an individual, group or scene in an image, which is the boundary between one object and another object, and between the object and the background.

Exemplarily, if the target image is a house image, and the target image includes chandeliers, beds, tables, windows, curtains, stairs and other related content about the house, the chandeliers, beds, tables, and windows can be extracted from the target image. , curtains, stairs, etc. contour edge lines. As an optional embodiment, edge detection algorithms such as Canny, Sobel, Prewitt, Roberts, Marr-Hildreth, etc. can be used to extract the contour edge lines of all contents in the target image.

In a specific implementation, the target image may be a screenshot image stolen by taking a screenshot. Specifically, the screenshot image is usually obtained by a user by taking a screenshot of an image displayed in an application program. When displaying images in an application, in order not to affect the user's normal browsing of images, the images are usually displayed on the background preset in the application, wherein the background is usually a solid color, such as a black background or a white background, which is exactly In this way, the screenshot image not only includes the screenshot image itself, but also includes the background when the screenshot image is displayed in the application.

Several specific examples are used to describe the screenshot image below: Referring to FIG. 3 , it is one of the screenshot images of a house image according to the present invention. In addition to the screenshot image 302 , the image also includes a background 304 ; 4, is the second screenshot image of a house image according to the present invention, in addition to the screen shot image 402, the image also includes a background 404; with reference to FIG. Third, in addition to the screen shot image 502 , the image also includes a background 504 .

Based on the above-mentioned feature that the screenshot image contains a background, if the target image is a screenshot image, the contour edge line extracted from the target image includes the background of the application program in addition to the content of the screenshot image itself. Taking the house image as an example, the contour edge lines extracted from the house image will not only contain the chandeliers, beds, tables, windows, curtains, stairs, etc. Outline edge lines. Exemplarily, referring to FIG. 6 , which is a schematic diagram of a contour edge line of a house image according to the present invention, the figure includes a contour edge line 602 corresponding to the background and contour edge lines of other contents in the house image.

Step 206: Determine an edge region in the target image based on the contour edge line.

The edge area refers to the image area determined based on the contour edge line in the target image.

In this embodiment of the present invention, after all contour edge lines in the target image are obtained, one or more edge regions may be determined in the target image based on the contour edge lines. Exemplarily, referring to FIG. 6 , the white image areas 604 in the figure all belong to the edge areas in the target image determined based on the contour edge lines 602 .

Step 208: Determine the color distribution of the edge region.

Step 210: When the color distribution condition satisfies a preset color distribution condition, identify the target image as a screenshot image; wherein the color distribution condition is based on the background corresponding setting when viewing the image in the application program.

The color distribution is used to represent the overall distribution of the color values of the pixels in the edge area, for example, a color to which the color values of the pixels in the edge area are biased, for example, biased toward black, or biased toward white.

Wherein, the color distribution condition is based on the corresponding setting of the background when viewing the image in the application program. Therefore, in the embodiment of the present invention, after determining one or more edge regions in the target image based on the contour edge line, each edge region is further determined. The color distribution of the area. If the color distribution of the edge area satisfies the preset color distribution conditions, it means that the target image includes the background when the image is displayed in the application, and the target image can be recognized as a screenshot image. Optionally, computer vision technology can be used to set the color distribution conditions based on the background when viewing the image in the application.

In the above image recognition method, the target image submitted to the application is acquired, and the contour edge line is extracted from the target image. After determining the edge area in the target image based on the contour edge line, the color distribution of the edge area is determined. If the color distribution If the condition satisfies the preset color distribution condition, the target image can be identified as a screenshot image. The color distribution condition is based on the corresponding setting of the background when viewing the image in the application. If the color distribution of the edge area satisfies the preset color distribution condition , indicating that the target image may be a screenshot image with a screenshot of the background. The embodiment of the present invention determines whether the target image is a screenshot image by automatically identifying whether the edge area of the target image satisfies the color distribution condition, thereby improving the recognition efficiency of the screenshot image.

On the basis of the above-mentioned embodiment, a modified embodiment of the above-mentioned embodiment is proposed. It should be noted here that, in order to simplify the description, only the difference from the above-mentioned embodiment is described in the modified embodiment.

In an exemplary embodiment, the step 206 of determining an edge region in the target image based on the contour edge line may include the following steps:

respectively acquiring at least two pixel coordinates of the contour edge line and the distance between the contour edge line and the edge of the target image;

Determine the angle of the contour edge line according to the pixel coordinates;

Determining a preset number of the contour edge lines whose angle is a preset angle and the distance from the edge of the target image is the smallest as the target contour edge line;

The target image between the target contour edge line and the edge of the target image is determined as an edge area.

In the specific implementation, the image displayed in the application program is usually rectangular, so if a screenshot of the image displayed in the application program is taken, the contour edge line corresponding to the background in the obtained screenshot image, if based on the horizontal line, its angle It should be a straight line of 90 degrees or 0 degrees, so in an optional example of the present invention, the preset angle may be set to 90 degrees or 0 degrees. In addition, the background portion in the screenshot image is located at the position closest to the edge of the image in the screenshot image, and more specifically, is located in the upper, lower, left and right positions in the screenshot image near the edge of the screenshot image, This is because when the screenshot image is displayed in the application, it is usually displayed in the middle of the background, so in an optional example of the present invention, the preset number can be set to 4.

In the embodiment of the present invention, when the contour edge line is extracted from the target image, at least two pixel coordinates of the contour edge line are obtained respectively. If the two pixel coordinates are both on the horizontal axis or the vertical axis of the coordinate system, it can be determined that The angle of the contour edge line is 90 degrees or 0 degrees, that is, the preset angle is satisfied, and the contour edge line is a straight line. If the angle of the contour edge lines is a preset angle, the distances between these contour edge lines and the edge of the target image are further determined, and a preset number of contour edge lines with the smallest distance from the edge of the target image, for example , determine the 4 contour edge lines with the smallest distance from the edge of the target image as the target contour edge line, and then, you can combine the target contour edge line and the edge of the target image. The target image, identified as the edge region.

It should be noted that the preset angle and the preset number can be set according to the actual situation when the image is displayed in the application. For example, if the target image is displayed obliquely in the application, the preset angle can be adjusted according to the inclination angle. Assuming that the left and right backgrounds in the application program are completely covered by the target image, the preset number may be set to 2, which is not limited in this embodiment of the present invention.

Illustratively, referring to FIG. 6 , the four dotted lines on the upper, lower, left and right in FIG. 6 are the edges of the target image. Among them, assuming that the contour edge line 602 in FIG. 6 is determined as the target contour edge line, then a rectangular frame can be obtained by combining the target contour edge line 602 and the dotted line (ie, the edge of the target image), and the target image in the rectangular frame is the edge. area 604.

In the above-mentioned exemplary embodiment, after the contour edge line is extracted from the target image, a preset number of which the angle is a preset angle and the distance from the edge of the target image is the smallest will be determined from the contour edge line. The contour edge line is combined with the edge of the target image to determine the edge area, which is used to determine whether the target image is a screenshot image based on the color distribution of the edge area, which can reduce the number of edge areas that need to be determined and reduce computing resources. consumption.

It should be noted that, in the embodiment of the present invention, the edge detection algorithm is used to extract the contour edge line from the target image. If the precision of the edge detection algorithm is not high, the extracted contour edge line will not be accurate enough, and the background in the screenshot image is different from the object image. The color contrast of the screenshot image is relatively large, so the contour edge line corresponding to the background in the screenshot image is very easy to be detected by the edge detection algorithm. In view of the above characteristics, the embodiment of the present invention can eliminate the need to use a high-precision edge detection algorithm, even for the target image. Some details cannot be detected accurately, and it does not affect the recognition of the target image.

In an exemplary embodiment, after the determining the angle of the contour edge line according to the pixel coordinates, the method may further include the following steps:

When there is no contour edge line whose angle is a preset angle, the target image is identified as a non-screenshot image.

In the embodiment of the present invention, if the coordinates of at least two pixels of the contour edge line are not both on the horizontal axis or the vertical axis of the coordinate system, it can be determined that the angle of the contour edge line is not 90 degrees or 0 degrees, that is, it does not meet the predetermined Set the angle, indicating that there is no contour edge line corresponding to the background in the target image, and it is not a straight line, then the target image can be identified as a non-screenshot image, that is, a compliant image.

In the above exemplary embodiment, when there is no contour edge line with a preset angle in the contour edge line of the target image, the target image can be directly recognized as a non-screenshot image, without the need to extract the contour edge line from the contour edge line. Determining a preset number of contour edge lines with the smallest distance from the edge of the target image improves the recognition efficiency of the target image.

In an exemplary embodiment, the step 208, determining the color distribution of the edge region, may include the following steps:

obtaining the pixel value of the pixel in the edge area;

The pixel value is used to calculate the mean and variance of the edge area; the mean and the variance are used to characterize the color distribution of the edge area.

Wherein, the mean value is the average value of the color values of the pixels in the image, and based on the mean value, the biased color of the image can be determined, for example, biased toward black or biased toward white. Variance reflects how far the color values of pixels in an image deviate from the mean.

In the embodiment of the present invention, the pixel values of the pixels of each edge region in the target image are obtained to calculate the mean and variance of the edge region based on the pixel values, so that the color distribution of the edge region can be determined according to the mean and variance. Of course, in addition to determining the color distribution of the edge region based on the mean and variance, the color distribution of the edge region can also be determined based on the mean square value, mean square error, and standard deviation of the edge region. This need not be restricted.

In the above exemplary embodiment, the color distribution of the edge region is determined according to the mean value and variance calculated from the pixel values of the pixels in the edge region, which is simple, accurate and efficient.

Specifically, the preset color distribution condition may be that the specified color is satisfied, and the specified color is the color of the specified image. In this case, when the color distribution condition satisfies the specified color and the target image is not the specified image, the target image is identified as a screenshot image.

In an exemplary embodiment, the specified image may be a floor plan, and the specified color is the color of the floor plan. In step 210, when the color distribution satisfies a preset color distribution condition, identify the target image as a screenshot. image, which can include the following steps:

When the mean value and the variance satisfy a preset color distribution condition corresponding to a specified color background, and the target image is not a floor plan, the target image is identified as a screenshot image.

The preset color distribution conditions include a preset mean and a preset variance corresponding to the background when the application is viewing an image. Therefore, if the mean and variance satisfy the preset mean and preset variance, the target image can be identified. for screenshots. Exemplarily, if the target image is in the RGB color mode, each pixel in the target image is assigned a pixel value in the range of 0 to 255, where 0 means black and 255 means white. Assuming that the mean value and variance std are calculated based on the pixels of the edge area, when mean>220 and std<10, it means that the edge area tends to be white, and the deviation of the overall pixel value is small, then the edge area can be determined as white area, when mean<5 and std<10, it means that the edge area is inclined to black, and the deviation of the overall pixel value is small, then the edge area can be determined as a black area. In this embodiment of the present invention, if it is determined that the edge area is a white area or a black area, the target image may be identified as a screenshot image.

Of course, if the target image is in CMYK color mode, HSB color mode, YUV color mode, etc., you can also convert the edge area to RGB color mode and then calculate the mean and variance to determine whether the target image is a screenshot image. Embodiments of the present invention This is not restricted.

In a specific implementation, the target image may be a house image, and the house image may also include a floor plan in addition to the house photo. Specifically, referring to FIG. 7 , it is a schematic diagram of a floor plan provided in an embodiment of the present invention, The floor plan refers to the plan space layout of the house, that is, the diagram that describes the use function, corresponding location and size of each independent space of the house. In practical applications, in order to be able to see the layout of the house more intuitively, most of the floor plans are white or white.

In an optional embodiment, the specified image is a floor plan, and the specified color is a color of the floor plan. For example, if the color of the floor plan is white, the specified color is white. In this embodiment of the present invention, considering that the target image may be a floor plan, if the mean and variance of the edge area satisfy the color distribution conditions set for the background of a non-specified color (for example, black), the target image can be identified as Screenshot image, if the mean and variance in the edge area satisfy the color distribution conditions set for the specified color (for example, white) background, it will be further determined whether the target image is a floor plan, specifically, if the target image both satisfies the specified color The background corresponding to the set color distribution conditions, and it is not a floor plan, the target image can be identified as a screenshot image.

In the above exemplary embodiment, if the mean and variance of the edge area satisfy the color distribution conditions set for the background of the specified color, indicating that the target image may be a special image, such as a floor plan, it can be determined whether the target image is a floor plan to further identify whether the target image is a screenshot image, avoid misrecognition, and improve the recognition accuracy.

In an exemplary embodiment, whether the target image is a floor plan is determined by:

Obtain the image information of the target image submitted in the application;

Whether the target image is a floor plan is determined based on the image information.

In a specific implementation, when a user submits an image in an application, image information about the image can also be submitted. For example, if the image is a floorplan, image information indicating that the image is a floorplan can be submitted in the application.

In the embodiment of the present invention, if the mean value and variance of the edge area satisfy the color distribution conditions set corresponding to the background of the specified color, for example, when the color distribution conditions corresponding to the white background are set, it means that the image may be a floor plan, and then obtain The image information of the target image is used to determine whether the target image is a floor plan based on the image information. If it is determined based on the image information that the target image is indeed a floor plan, the target image can be identified as a non-screenshot image, that is, a compliant image. If the image information determines that the target image is not a floor plan, the target image can be identified as a screenshot image, that is, an illegal image.

In the above exemplary embodiment, the image information of the target image submitted by the user is acquired to identify whether the target image is a screenshot image in combination with preset color distribution conditions and image information, which can improve the recognition accuracy of the target image.

In an exemplary embodiment, after the determining the color distribution of the edge region, the method may further include the following steps:

When the color distribution condition does not meet the preset color distribution condition, the target image is identified as a non-screenshot image.

In the embodiment of the present invention, after determining one or more edge regions in the target image, the color distribution of each edge region is further determined. Without including the context in which the image is displayed in the application, the target image can be identified as a non-screenshot image, i.e. a compliant image.

Exemplarily, it is assumed that the mean value and the variance std are calculated based on the pixels of the edge area, that is, when the mean>220 and std<10 are not satisfied, and the mean<5 and std<10 are not satisfied, it means that the color distribution of the edge area is not good. If the preset color distribution conditions are met, the target image can be recognized as a non-screenshot image.

In the above exemplary embodiment, if the edge region of the target image does not satisfy the color distribution condition, indicating that the target image is not a screenshot image, the target image can be identified as a non-screenshot image, which improves the recognition efficiency of the target image.

In an exemplary embodiment, after the target image is identified as a screenshot image when the color distribution condition satisfies a preset color distribution condition, the method may further include the following steps:

If the target image is identified as a screenshot image, publishing a post containing the target image in the application is prohibited.

In a specific implementation, the target image may be an image in a post submitted by a user to the application, for example, a house image in a post on a house transaction. If the target image is identified as a screenshot image, it means that the image in the post may be an existing image. Images after illegal secondary use of images, in order to ensure the property safety and life safety of all users in the application, posts containing screenshots will not be allowed to be published in the application.

In the above exemplary embodiment, if it is recognized that the image in the post submitted by the user to the application is a screenshot image, the post will be prohibited from being published in the application, which improves the review efficiency of the post, especially when the number of posts is large. Effectively reduce the release of false information on the application, improve the accuracy of the posts displayed in the application, avoid unnecessary trouble for users caused by false information posts, users can use the application with confidence, and improve the user's trust. .

In order to make those skilled in the art better understand the embodiments of the present invention, a specific example is used for description below. Referring to FIG. 8, it is a flowchart of an image recognition provided by an embodiment of the present invention, including the following steps:

Step 802: Acquire a target image.

Step 804, using the canny edge detection algorithm to extract all contour edge lines in the target image;

Step 806: Extract the contour edge line (straight line) with an angle of 0 degrees or 90 degrees through the pixel coordinates on the same contour edge line, and further screen out the contour edge line closest to the edge of the target image, that is, the distance from the target image. The top, bottom, left, and right edges of the image are the closest four contour edge lines. If the target image includes the target contour edge line whose angle and distance both meet the conditions, go to step 808 , and if the target image does not include the target contour edge line whose angle and distance both meet the conditions, go to step 816 .

Step 808: Combine the edge of the target contour and the edge of the target image into a rectangular frame, the target image in the rectangular frame is the edge area, and calculate the mean and variance std of the pixels in the edge area.

Step 810, when mean>220 and std<10, determine the edge area as a white area, when mean<5 and std<10, determine the edge area as a black area, if the edge area is a white area, then execute step 812 , and if the edge area is a black area, step 814 is executed. Among them, if the mean and variance std of the edge area do not satisfy the color distribution conditions of mean>220 and std<10, or mean<5 and std<10, that is, the edge area is neither black nor white, then perform the steps 816.

Step 812: Considering that if the target image is a floor plan, the edge area may be identified as a white area, which will cause false calls. Therefore, when the edge area of the white area is extracted from the target image, it is necessary to further obtain the image provided by the user. information, determine whether the target image is a floor plan, if yes, go to step 816, if not, go to step 814.

Step 814: Determine the screenshot image.

Step 816: Determine the image as a non-screenshot image.

By applying the embodiments of the present invention, when a user publishes a post in an application program, an image in the post can be identified as a target image, and whether it is a screenshot image stolen by a screenshot method can be ensured under the premise that deployment resources are less expensive. , for high recall of illegally used screenshot images, it has the characteristics of stable features, high recognition accuracy, and high robustness, which effectively saves the cost of human review.

It should be noted that, for the sake of simple description, the method embodiments are described as a series of action combinations, but those skilled in the art should know that the embodiments of the present invention are not limited by the described action sequences, because According to embodiments of the present invention, certain steps may be performed in other sequences or simultaneously. Secondly, those skilled in the art should also know that the embodiments described in the specification are all preferred embodiments, and the actions involved are not necessarily required by the embodiments of the present invention.

Referring to FIG. 9, a structural block diagram of an image recognition apparatus provided in an embodiment of the present invention is shown, which may specifically include the following modules:

a target image acquisition module 902, configured to acquire the target image submitted to the application;

an outline edge line extraction module 904, used for extracting outline edge lines from the target image;

an edge region determination module 906, configured to determine an edge region in the target image based on the contour edge line;

a color distribution determination module 908, configured to determine the color distribution of the edge region;

The image recognition module 910 is configured to recognize the target image as a screenshot image when the color distribution condition satisfies a preset color distribution condition; wherein, the color distribution condition is based on the image when viewing the image in the application program. Background corresponding settings.

In an exemplary embodiment, the edge region determination module 906 is configured to obtain at least two pixel coordinates of the contour edge line and the distance between the contour edge line and the edge of the target image respectively; The angle of the contour edge line is determined according to the pixel coordinates; the preset number of contour edge lines whose angle is a preset angle and whose distance from the edge of the target image is the smallest are determined as the target contour edge line; the target image between the target contour edge line and the edge of the target image is determined as an edge area.

In an exemplary embodiment, the edge area determination module 906 is configured to identify the target image as a non-screenshot image when there is no contour edge line whose angle is a preset angle.

In an exemplary embodiment, the color distribution determination module 908 is configured to obtain pixel values of pixels in the edge area; use the pixel values to calculate the mean and variance of the edge area; the mean and all The variance is used to characterize the color distribution of the edge region.

In an exemplary embodiment, the image recognition module 910 is configured to, when the mean value and the variance satisfy the preset color distribution conditions set for the background of the specified color, and the target image is not a floor plan , identifying the target image as a screenshot image.

In an exemplary embodiment, the image recognition module 910 is configured to acquire image information of the target image submitted in the application; determine whether the target image is a floor plan based on the image information.

In an exemplary embodiment, the image recognition module 910 is configured to recognize the target image as a non-screenshot image when the color distribution does not meet a preset color distribution condition.

In an exemplary embodiment, the apparatus further includes a publishing prohibiting module, configured to prohibit publishing a post containing the target image in the application program if the target image is identified as a screenshot image.

In the embodiment of the present invention, the target image submitted to the application is acquired, the contour edge line is extracted from the target image, and after determining the edge area in the target image based on the contour edge line, the color distribution of the edge area is determined, if the color distribution If the condition satisfies the preset color distribution condition, the target image can be identified as a screenshot image. The color distribution condition is based on the corresponding setting of the background when viewing the image in the application. If the color distribution of the edge area satisfies the preset color distribution condition , indicating that the target image may be a screenshot image with a screenshot of the background. The embodiment of the present invention determines whether the target image is a screenshot image by automatically identifying whether the edge area of the target image satisfies the color distribution condition, thereby improving the recognition efficiency of the screenshot image.

As for the apparatus embodiment, since it is basically similar to the method embodiment, the description is relatively simple, and reference may be made to the partial description of the method embodiment for related parts.

Preferably, an embodiment of the present invention further provides an electronic device, comprising: a processor, a memory, a computer program stored in the memory and running on the processor, and the computer program is executed by the processor to implement the above image recognition method Each process of the embodiment can achieve the same technical effect, and to avoid repetition, it will not be repeated here.

Embodiments of the present invention also provide a computer-readable storage medium, where a computer program is stored on the computer-readable storage medium, and when the computer program is executed by a processor, each process of the above image recognition method embodiment is implemented, and the same technology can be achieved. The effect, in order to avoid repetition, is not repeated here. The computer-readable storage medium is, for example, a read-only memory (Read-Only Memory, ROM for short), a random access memory (Random Access Memory, RAM for short), a magnetic disk, or an optical disk.

An embodiment of the present invention provides a computer program product, the program product is stored in a storage medium, and the program product is executed by at least one processor to implement the various processes in the above image recognition method embodiments, and can achieve the same technical effect , in order to avoid repetition, it will not be repeated here.

It should be noted that, herein, the terms "comprising", "comprising" or any other variation thereof are intended to encompass non-exclusive inclusion, such that a process, method, article or device comprising a series of elements includes not only those elements, It also includes other elements not expressly listed or inherent to such a process, method, article or apparatus. Without further limitation, an element qualified by the phrase "comprising a..." does not preclude the presence of additional identical elements in a process, method, article or apparatus that includes the element.

From the description of the above embodiments, those skilled in the art can clearly understand that the method of the above embodiment can be implemented by means of software plus a necessary general hardware platform, and of course can also be implemented by hardware, but in many cases the former is better implementation. Based on this understanding, the technical solutions of the present invention can be embodied in the form of software products in essence or the parts that make contributions to the prior art, and the computer software products are stored in a storage medium (such as ROM/RAM, magnetic disk, CD), including several instructions to make a terminal (which may be a mobile phone, a computer, a server, an air conditioner, or a network device, etc.) execute the methods described in the various embodiments of the present invention.

The embodiments of the present invention have been described above in conjunction with the accompanying drawings, but the present invention is not limited to the above-mentioned specific embodiments, which are merely illustrative rather than restrictive. Under the inspiration of the present invention, without departing from the spirit of the present invention and the scope protected by the claims, many forms can be made, which all belong to the protection of the present invention.

Those skilled in the art can realize that the units and algorithm steps of each example described in conjunction with the embodiments disclosed in the embodiments of the present invention can be implemented by electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Skilled artisans may implement the described functionality using different methods for each particular application, but such implementations should not be considered beyond the scope of the present invention.

Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the above-described systems, devices and units may refer to the corresponding processes in the foregoing method embodiments, which will not be repeated here.

In the embodiments provided in this application, it should be understood that the disclosed apparatus and method may be implemented in other manners. For example, the apparatus embodiments described above are only illustrative. For example, the division of the units is only a logical function division. In actual implementation, there may be other division methods. For example, multiple units or components may be combined or Can be integrated into another system, or some features can be ignored, or not implemented. On the other hand, the shown or discussed mutual coupling or direct coupling or communication connection may be through some interfaces, indirect coupling or communication connection of devices or units, and may be in electrical, mechanical or other forms.

The units described as separate components may or may not be physically separated, and components displayed as units may or may not be physical units, that is, may be located in one place, or may be distributed to multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution in this embodiment.

In addition, each functional unit in each embodiment of the present invention may be integrated into one processing unit, or each unit may exist physically alone, or two or more units may be integrated into one unit.

The functions, if implemented in the form of software functional units and sold or used as independent products, may be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention can be embodied in the form of a software product in essence, or the part that contributes to the prior art or the part of the technical solution. The computer software product is stored in a storage medium, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in the various embodiments of the present invention. The aforementioned storage medium includes: a U disk, a removable hard disk, a ROM, a RAM, a magnetic disk, or an optical disk and other mediums that can store program codes.

The above are only specific embodiments of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art can easily think of changes or substitutions within the technical scope disclosed by the present invention. should be included within the protection scope of the present invention. Therefore, the protection scope of the present invention should be subject to the protection scope of the claims.

Claims (11)

1. An image recognition method, characterized in that the method comprises:

acquiring a target image submitted to an application program;

extracting contour edge lines from the target image;

determining an edge region in the target image based on the contour edge line;

determining the color distribution condition of the edge area;

when the color distribution condition meets a preset color distribution condition, identifying the target image as a screen shot image; wherein the color distribution condition is based on a background correspondence setting when an image is viewed in the application program.

2. The method of claim 1, wherein determining an edge region in the target image based on the contour edge line comprises:

respectively acquiring at least two pixel coordinates of the contour edge line and a distance between the contour edge line and the edge of the target image;

determining the angle of the contour edge line according to the pixel coordinates;

determining a preset number of the contour edge lines with the angle as a preset angle and the minimum distance between the contour edge lines and the edge of the target image as target contour edge lines;

and determining the target image between the target contour edge line and the edge of the target image as an edge area.

3. The method of claim 2, wherein after said determining the angle of the contour edge line from the pixel coordinates, the method further comprises:

and when the contour edge line with the preset angle does not exist, identifying the target image as a non-screen capture image.

4. The method of claim 1, wherein determining the color distribution of the edge region comprises:

acquiring a pixel value of a pixel in the edge area;

calculating the mean and variance of the edge region by using the pixel values; the mean value and the variance are used for characterizing the color distribution condition of the edge area.

5. The method according to claim 4, wherein the identifying the target image as a screenshot image when the color distribution condition satisfies a preset color distribution condition comprises:

and when the mean value and the variance meet preset color distribution conditions correspondingly set for a background of a specified color and the target image is not a house-type image, identifying the target image as a screen shot image.

6. The method of claim 5, wherein determining whether the target image is a house type map is performed by:

acquiring image information of the target image submitted in the application program;

determining whether the target image is a house type map based on the image information.

7. The method according to claim 1, wherein after the determining the color distribution of the edge region, the method further comprises:

and when the color distribution condition does not meet the preset color distribution condition, identifying the target image as a non-screenshot image.

8. The method according to claim 1, wherein after the target image is identified as a screenshot image when the color distribution condition satisfies a preset color distribution condition, the method further comprises:

and if the target image is identified as a screen capture image, prohibiting a post containing the target image from being issued in the application program.

9. An image recognition apparatus, characterized in that the apparatus comprises:

the target image acquisition module is used for acquiring a target image submitted to an application program;

the contour edge line extraction module is used for extracting contour edge lines from the target image;

an edge region determining module, configured to determine an edge region in the target image based on the contour edge line;

the color distribution condition determining module is used for determining the color distribution condition of the edge area;

the image identification module is used for identifying the target image as a screen capture image when the color distribution condition meets a preset color distribution condition; wherein the color distribution condition is based on a background correspondence setting when an image is viewed in the application program.

10. An electronic device, comprising a processor, a communication interface, a memory and a communication bus, wherein the processor, the communication interface and the memory communicate with each other via the communication bus;

the memory is used for storing a computer program;

the processor, when executing a program stored on the memory, implementing the method of any of claims 1 to 8.

11. One or more computer-readable media having instructions stored thereon that, when executed by one or more processors, cause the processors to perform the method of any of claims 1-8.

Images