CN102867321A - Glasses virtual try-on interactive service system and method - Google Patents

Abstract

The invention relates to a virtual glasses try-on interactive service system and method taking a real person as a service object. The method comprises the following steps: the human face is positioned through a frame in the picture, and a corresponding human face image is obtained, so that a plurality of first feature point positions are obtained. And sampling a plurality of first feature information around each first feature point. And using picture comparison to judge whether the human face has dynamic movement, and obtaining a second human face image through the next picture. And combining the method of searching the contrast range and tracking the feature information to obtain a plurality of second feature points in the second face image, thereby obtaining a plurality of second feature point positioning information. And then, judging the position, the moving state and the scaling of the face by using the corresponding position information difference of the first face image and the second face image, and further calculating the positions of the plurality of second characteristic points. And synthesizing a preset glasses model at the positions of the second characteristic points.

Description

Glasses virtual try-on interactive service system and method

technical field

The present invention relates to a glasses three-dimensional (3D) virtual try-on interactive service system and method that provides real images, and in particular to a glasses 3D try-on interactive service system and method used in an e-commerce interactive system.

Background technique

With the vigorous development of e-commerce, more and more consumers are increasingly relying on the e-commerce interactive platform system to choose their favorite goods and commodities. The combined display pictures of products and models, and electronic product try-on systems and software are also attracting more and more attention from consumers, which in turn arouses consumers' desire to buy. The glasses try-on system that provides real-life images is one of the try-on systems that have attracted the attention and popularity of consumers. Users can find their favorite and suitable glasses from tens of thousands of glasses products as long as they use their own photos. .

However, most of the traditional glasses try-on systems are two-dimensional (2D) planes, and users can only watch the frontal image of the glasses they are trying on, but cannot watch the side images of themselves trying on the glasses from the left and right sides. Moreover, the traditional glasses try-on system cannot properly synthesize the glasses with the consumer's face as the consumer's face moves and rotates. Therefore, the synthesized image is often too abrupt or unreal.

Therefore, in view of the lack of an effective and economic mechanism to solve the traditional problems in the traditional glasses try-on system and method, it is urgent to propose a novel glasses virtual try-on interactive service system and method, which can solve the above problems through accurate simulation and calculation. The problem.

Contents of the invention

In order to solve the above-mentioned problems, the present invention provides an interactive service system and method for glasses virtual try-on with live images. After precise simulation and calculation, it solves the problem that the glasses and the face of the consumer cannot be matched when the face of the consumer moves and rotates. Proper synthesis, resulting in the problem that the synthetic image is too abrupt or unreal.

According to a specific embodiment, the present invention provides a virtual glasses try-on interactive service method for real-person images, which is characterized by: locating the face through the frame in the screen, and obtaining the first face image; The eyes of the image obtain a plurality of first feature points; sample a plurality of first feature information around each first feature point, and store the plurality of first feature information; use screen comparison to determine whether the face has dynamic movement, And obtain the second face image through the next frame; obtain a plurality of second feature points in the second face image by combining the search comparison range and tracking feature information method, and then obtain the positioning information of a plurality of second feature points; Comparing the relative position information difference between the first human face image and the second human face image to judge the position, moving state and zoom ratio of the human face, and then calculate the positions of the plurality of second feature points; The assumed glasses model is synthesized at the positions of the plurality of second feature points.

According to this specific embodiment, the present invention provides a virtual glasses try-on interactive service system for real-person images, which is characterized by: an image capture unit that locates a face through a frame in the screen and obtains a first face image; A processing unit, coupled to the image capture unit, obtains a plurality of first feature points for the eyes of the first human face image, and designs a sampling pattern around each first feature point to obtain a plurality of favorable first features information, and store the plurality of first feature information, and use the screen comparison to judge whether the face has dynamic movement, and obtain the second face image through the next screen, combine the search comparison range and tracking feature information method and Obtaining a plurality of second feature points in the second face image, and then obtaining location information of a plurality of second feature points; an analysis unit, coupled to the processing unit, for comparing the first face image with the second face The relative position information difference of the image is used to judge the position, moving state and zoom ratio of the face, and then calculate the positions of the plurality of second feature points; The glasses model is synthesized at the positions of the plurality of second feature points. Follow up by analogy with this method to obtain the positions of the dynamic third and fourth feature points.

According to another specific embodiment, the present invention provides a contact lens virtual try-on interactive service method, which is characterized in that it includes: locating a human face through a frame in the screen, and obtaining a first human face image; The eye pupil of the image is obtained by obtaining a plurality of first feature points; sampling a plurality of first feature information around each first feature point, and storing the plurality of first feature information; and synthesizing a preset contact lens model on the The positions of the plurality of second feature points. Follow up by analogy with this method to obtain the positions of the dynamic third and fourth feature points.

In order to provide a more in-depth description of the present invention, the following illustrations, figure number descriptions and detailed descriptions of the invention are used, hoping to be helpful to the examination work.

Description of drawings

Fig. 1A and Fig. 1B are a kind of glasses virtual try-on interactive service method according to a specific embodiment of the present invention;

Fig. 2A～Fig. 2D are the operation means according to Fig. 1 method;

3A to 3C show the tilt, rotation and movement angles of the human face;

Fig. 4 is a schematic diagram of the state of glasses synthetic human face;

5A to 5C are the manufacturing method of 3D glasses according to the specific embodiment of the present invention;

6A-6B are a contact lens virtual try-on interactive service method according to another embodiment of the present invention;

Fig. 7A～Fig. 7B are the operation means according to Fig. 6 method;

Fig. 8 is a virtual try-on interactive service system for glasses according to a specific embodiment of the present invention.

reference sign

Steps s101 to s107

Steps s601-s607

81 image capture unit

82 processing units

83 analytical units

84 synthesis units

85 glasses database

Detailed ways

Use the following diagrams to illustrate the detailed structure and connections of this creation.

Fig. 1 is a real-person image according to a specific embodiment of the present invention, a kind of glasses virtual try-on interactive service method, which is characterized by: positioning the face through the frame in the screen, and obtaining the first face image (step s101), For example, as shown in FIG. 2A , the present embodiment displays a dotted line frame on the screen and requires the front of the user's face to enter the frame to match the size of the dotted line frame and align the eyes with the horizontal line to locate the user's face, and The first human face image is taken. Next, obtain a plurality of first feature points (step s102) at the eye features of the first human face image, such as the corners of the eyes on both sides (step s102). The sides of the mouth. For example, as shown with a cross in Figure 2B, in this specific embodiment, the corner points on both sides of the left and right eyes of the first human face image and the points on both sides of the corner of the mouth can be manually selected to obtain one or more first features respectively. point. In addition, in another specific embodiment, the face recognition method can also be used to allow the program to automatically capture feature points. In addition, after this step (s102), this embodiment can further judge whether the first human face image conforms to the human face logic according to the plurality of first feature points, and if not, then reacquire the first human face image or the plurality of first feature points. If so, proceed to the next step, apply a search operation to determine whether the plurality of first feature points are located in the frame, if not, then re-acquire the first human face image or the plurality of first feature points, if so , proceed to the next step. Among them, the way to judge whether the position of the feature points conforms to the logic of the face includes whether the corner points on both sides of the left (right) eye are in the left (right) half of the dotted line frame, whether the inner and outer corners of the same eye are separated by a certain distance, whether the inner and outer corners of the same eye Whether the height difference between the corners of the eyes or the left and right corners of the mouth is too large, etc. If one of the above-mentioned limiting conditions is not met, it is required to obtain the first feature point or the first human face image again.

Next, design a sampling pattern around each first feature point to obtain favorable multiple first feature information, and store the multiple first feature information (step s103), and the features include simultaneously storing the multiple first feature information A point distance between the point and the plurality of first feature points. In terms of sampling mode, the feature information is pixel color information, and the pixel color information is radiated in m directions from each feature point, and n pixels are obtained in each direction as color information, and the m and n are positive integers. Alternatively, the pixel color information is obtained from m directions radiated by each feature point in a semicircle, and n pixels are obtained in each direction as color information, and the m and n are positive integers and the semicircle Cover at least the corner of one eye. As shown in FIG. 2C , in this specific embodiment, the color information of adjacent pixels of feature points is captured by the method of sampling mode, and the point spacing between each feature point is recorded at the same time. For example, the feature point extends outward in 8 directions in the form of semicircle radiation diffusion, and takes 7 points and a total of 56 color information as the feature information of the feature point, and the semicircle completely covers at least the corner of the eye. Or, as shown in Figure 2D, in another specific embodiment, the sampling mode extends outward in 8 directions from the feature point in a radial diffusion manner, and takes 7 points and a total of 56 color information as the feature point feature information.

Next, obtain the second human face image in the next frame, and use frame comparison to determine whether the human face has dynamic movement (step s104 ). Wherein, within the time interval, compare the face in the frame with the face in the next frame, and dynamically track whether the plurality of first feature points have moving tracks. For example, in this specific embodiment, this picture is compared with the next picture pixel in a subtractive manner to the object moving within this time interval, if there are obvious moving traces near the feature points, it means that the face has moved during this time, then Carry out subsequent steps and operations. Conversely, if there is no movement on the face, no follow-up processing will be performed. In another specific embodiment, objects moving within the time interval may also be compared in a manner of subtracting a background image from the frame. Alternatively, the number of white dots (moving dots) in the feature information of the pixel difference between the front and back images determines the degree of movement in the current image. If the number of white dots is large, it means that the face is moving, and the subsequent steps and calculations are performed. On the contrary, if there is no white point near the feature point, it means that the face has no obvious moving traces, and no follow-up tracking process will be performed.

Then, the noise filtering method is applied to filter out the noise in the next picture, so as to avoid the noise interference in the subsequent comparison process and increase the error rate, and the noise filtering method is one of the Gaussian blur method, the median method and the mean method one.

When there are moving traces mentioned above, it will further combine the search and comparison range and track feature information, and then obtain a plurality of second feature points in the second face image, and then obtain a plurality of second feature information (step s105), and compare the The relative position information difference between the first human face image and the second human face image is used to determine the position, moving state and scaling ratio of the human face, and then calculate the positions of the plurality of second feature points (step s106 ). Wherein, at least a comparison range is preset to compare the plurality of first characteristic information and the plurality of second characteristic information, and take a plurality of error values between the plurality of first characteristic information and the plurality of second characteristics, And sort the plurality of error values, and obtain the first i smallest error values, and then obtain the positions of the plurality of second feature points. In this specific embodiment, there are different decision methods according to two different states: (1) search state: when the feature points are just selected and start detection for the first time or when the tracking fails, it will enter this state. In this state, the range of comparison is static, and the comparison is only limited to the adjacent range of the feature point selected by the user for the first time, which means that the user is forced to face the position of the dotted frame to enter the comparison range. (2) Tracking state: when the feature points are successfully compared in the previous screen, it is in the tracking state. In this state, the compared area is the adjacent area of the feature points compared in the previous screen, which means this area It is dynamic and will move with the feature points currently being chased. And within the scope of the comparison, it will include sampling N pixels according to the designed sampling mode, and each pixel will use the sampling mode to obtain 56 points adjacent to the pixel point, and the RGB and YCbCr color information of these points will be compared with the initial The recorded first feature information is compared, the error value of the two is Error 1 ~ Error N, sort the N values, take the first i pixels with the smallest error as candidate points, for example, 10, and do another grouping Remove outliers, and finally average these concentrated pixel value coordinates, and the result is the final tracking result. If the Error values found in the N points are all too large, it means that the user’s movement has exceeded the tracking range, or there are additional occluders appearing in front of the feature points. At this time, it is determined that the tracking has failed, and no further follow-up will be performed. operation.

In addition, this example can further perform geometric calculations, for example, according to the positions of the plurality of first feature points and the plurality of second feature points, calculate the inclination of the face with the slope, and calculate the slope of the human face according to the distance between the plurality of first feature points. The distance between the distance and the plurality of second feature points, the distance ratio of the face is calculated by the length change, and the ratio of the human face is calculated by the ratio of the first feature point and the second feature point. Face rotation and pitch angles. If the inclination, the distance ratio, the rotation angle or the pitch angle of the human face exceeds a preset allowable value, the second human face image is reacquired. As shown in FIG. 3A , according to the coordinates of the corners of the two eyes on both sides of the first face image are located on the horizontal axis, and the coordinates of the corners of the eyes on both sides of the second face image are upward angles, therefore, The inclination of the face can be calculated using the slope. As shown in FIG. 3B , according to the distance D1 between the eye corners of the first human face image and the distance D2 between the two eye corners of the second human face image, the distance ratio of the human face is calculated according to the length change. As shown in Figure 3C, according to the ratio of the two eye corners of the two eyes in the first face image and the ratio of the two eye corners in the second face image to the ratio of the two eyes in the first face image The ratio of the side corners of the eyes is 1.1:1 or 1.2:1, the rotation angle of the face is calculated by the ratio change and according to the two eyes and the mouth in the first face image are on the horizontal axis, and the two eyes in the second face image Compared with the mouth, the ratio of the two eyes to the mouth in the first human face image is 1:1.1, and the pitch angle of the human face is calculated according to the ratio change.

Next, the preset glasses model is extracted from the preset glasses database to synthesize the positions of the plurality of second feature points (step s107), and according to the size of the face, the plurality of first feature points and the error value, The glasses model can be scaled and rotated, and then synthesized to the human face appropriately, and the glasses database is used to store the data of the glasses model. As shown in Figure 4, the three-dimensional space information calculated by the above steps can intelligently rotate the glasses model to a position that conforms to the current user's movement, and because the faces are different in size, it is necessary to match the first feature of the majority The point coordinates scale and rotate the size of the glasses and synthesize them on the face of the person. After this step, the method can end, or continue to the next search and trace step. In this specific embodiment, the glasses model is a three-dimensional (3D) glasses model, and its manufacturing method can be divided into three directions: the front, the front left side, and the right side of a physical glasses through an imaging device (for example, a digital camera). Obtain the plane image of the real glasses (as shown in FIGS. 5A and 5B ), but is not limited thereto. Alternatively, the real glasses are rotated by plus or minus 90 degrees to obtain three plane images of the front, front left and right sides of the real glasses. Then, the three obtained plane images are combined into the three-dimensional glasses model (as shown in FIG. 5C ) by applying image synthesis software and hardware. After the 3D glasses model is completed, various parameters (such as rotation parameters, color parameters, scale parameters and transparency parameters, etc.) can also be applied to adjust the position and color of the 3D glasses, or to apply options (such as, Select the front view or the left and right side views) to enjoy the three-dimensional glasses model (as shown in Figure 5C). This specific embodiment only considers the manufacture of a three-dimensional frame without spectacle lenses. However, those skilled in the art should understand that the present invention can also be used to manufacture spectacle models with spectacle lenses.

Fig. 6 is a contact lens virtual try-on interactive service method according to another specific embodiment of the present invention. The difference between this specific embodiment and the above specific embodiment is that this embodiment uses the eye pupil as a feature point and combines the sampling mode to obtain feature information, and is suitable for virtual simulation of trying on contact lenses, while other methods and calculations are the same as the above specific implementation Therefore, only the details of these differences are described in detail in this example, and the steps and methods similar to the above-mentioned specific embodiments are not repeated here. The characteristics of the 3D simulation method in this embodiment include: locating a human face through a frame in the frame, and obtaining a first human face image (step s601 ). A plurality of first feature points (step s602) are obtained at the eye pupil feature of the first human face image, and the plurality of first feature points include the pupils of the left and right eyes and both sides of the corners of the mouth, for example, as shown in Figure 7A with a cross As shown, in this specific embodiment, the pupil points of the left and right eyes and points on both sides of the corner of the mouth of the first human face image are manually selected to obtain one or more first feature points respectively. Sampling a plurality of first feature information around each first feature point, and storing the plurality of first feature information (step s603). In this specific embodiment, the way to judge whether the position of the feature points conforms to the logic of the face includes whether the pupil of the left (right) eye is in the left (right) half of the dotted line frame, whether the pupil of the same eye is separated from the corners of the eyes on both sides by a certain distance, whether the pupil of the eye Whether the height gap between the pupils and the left and right corners of the mouth is too large. If one of the above-mentioned limiting conditions is not met, it is required to obtain the first feature point or the first human face image again.

Next, obtain the second human face image in the next frame, and judge whether the human face has dynamic movement (step s604), then search for a comparison range and track feature information, and then obtain multiple first human face images in the second human face image two feature points, and then obtain a plurality of second feature information (step s605), and the feature includes storing the point distance between the plurality of first feature points at the same time. As shown in Figure 7B, the sampling method in this specific embodiment captures the color information of adjacent pixels of the feature points, and records the point spacing between each feature point at the same time, for example, the way of radial diffusion from the feature points extends outward to 8 direction, and take 7 points and a total of 56 color information as the feature information of this feature point. Comparing the first human face image and the second human face image to determine the position, moving state and scaling of the human face, and then calculate the positions of the plurality of second feature points (step s606), for example, according to the The coordinates of the pupil point in the first human face image are located on the horizontal axis, while the coordinates of the pupil point in the second human face image are the upward angle, therefore, the inclination of the human face can be calculated by using the slope. For example, according to the distance between the two pupils in the first human face image and the distance between the two pupils in the second human face image, the distance ratio of the human face is calculated according to the length change. For example, according to the ratio of the pupils of the two eyes in the first human face image to the pupils of the two eyes in the second human face image, the rotation angle of the face is calculated according to the ratio of the pupils of the two eyes in the first human face image and the ratio of the pupils of the two eyes in the first human face image to The ratio between the mouth and the pupils of the two eyes and the mouth in the second face image is compared with the ratio between the eyes and the mouth in the first face image, and the rotation angle and the pitch angle of the face are calculated according to the proportional changes. If the inclination, distance ratio, rotation angle or pitch angle of the face exceeds a preset allowable value, the second face image is reacquired. And synthesizing the preset contact lens model at the positions of the plurality of second feature points (step 607), and scaling and rotating the lens model according to the size of the face and the plurality of first feature points, and then suitable synthesized to the face. After this step, the method can end, or continue to the next search and trace step.

Fig. 8 is an interactive service system for glasses virtual try-on according to a specific embodiment of the present invention. The system features include: an acquisition unit 81 , a processing unit 82 , an analysis unit 83 and a synthesis unit 84 . The capture unit 81 locates the face through the frame in the frame, and obtains the first face image, and the capture unit 81 can be a photographing device, such as a network camera. The processing unit 82, coupled to the extraction unit 81, obtains a plurality of first feature points at the eye features of the first human face image, such as the corners of the eyes on both sides, and samples around each first feature point according to the sampling mode. A plurality of first feature information, and store the plurality of first feature information, and obtain the second face image in the next frame, and judge whether the face has dynamic movement, and obtain the second face image. A second feature point, and then obtain a plurality of second feature information, and a plurality of first feature points include but not limited to the corners of the eyes on both sides of the left and right eyes and both sides of the corner of the mouth. And the processing unit 82 is characterized by judging whether the first human face image conforms to the human face logic according to the plurality of first feature points, and reacquiring the first human face image if not. If match, then proceed to the next step, the processing unit 82 features include applied search operations to judge whether the plurality of first feature points are located in the frame, if not, then re-acquire the first human face image, if so, Then proceed to the next step. In the above sampling mode, the characteristic information is pixel color information, and the processing unit 82 radiates m directions from each characteristic point, and obtains n pixel points in each direction as color information, wherein the m and n are positive integers. Alternatively, the processing unit 82 radiates m directions from each of the feature points in a semicircle, and obtains n pixels in each direction as color information, and the semicircle covers at least the corner of one eye, wherein the m and n is a positive integer.

The analysis unit 83, coupled to the processing unit 82, is used to compare the difference in relative position information between the first human face image and the second human face image to determine the position, moving state and scaling of the human face, and then Calculate the positions of the plurality of second feature points. The analysis unit 83 compares the face in the frame with the face in the next frame within a time interval, and tracks whether the plurality of first feature points have moving tracks, and applies a noise filtering method to filter out the faces in the next frame. noise, and the noise filtering method is one of Gaussian blur method, median method and mean method. The analysis unit is also used to preset at least one comparison range to compare the plurality of first characteristic information and the plurality of second characteristic information, and take a complex number between the plurality of first characteristic information and the plurality of second characteristics error values, and sort the plurality of error values, and obtain the first i minimum error values, and then obtain the positions of the plurality of second feature points, and the above i is a positive integer. The analysis unit calculates the inclination of the human face according to the positions of the plurality of first feature points and the plurality of second feature points, and calculates the slope of the face according to the distance between the plurality of first feature points and the plurality of second feature points. the distance between the points, calculate the distance ratio of the human face with the length change and calculate the rotation angle and pitch angle of the human face with the proportional change according to the ratio of the plurality of first feature points to the plurality of second feature points, and If the inclination, distance ratio, rotation angle or pitch angle of the face exceeds a preset allowable value, the second face image is acquired again. The synthesis unit 84 is coupled to the analysis unit 83, extracts a preset glasses model from the preset glasses database 85 and synthesizes the preset glasses model at the positions of the plurality of second feature points, and the synthesis unit according to The size of the human face and the plurality of first feature points are used to scale and rotate the glasses model, and then synthesized to the human face. The glasses try-on virtual simulation system in this example can also be applied to the contact lens virtual try-on interactive service system through appropriate modification operations and steps.

The above is only the best embodiment of the present invention, and it cannot be used to limit the scope of the present invention. Without departing from the spirit and essence of the present invention, those skilled in the art should be able to implement the present invention according to the present invention. Various corresponding changes and modifications are made, but these corresponding changes and modifications should belong to the protection scope of the appended claims of the present invention.

Claims (41)

1. A virtual glasses try-on interactive service method, characterized in that, Locating a human face through a frame in a frame, and obtaining a first human face image; At the characteristic part of the eye, a plurality of first feature points and feature information are obtained according to the sampling mode, and the distance between the plurality of first feature information and the plurality of first feature points is stored; Using the dynamic image to judge and search and track the feature information to obtain a second face image in the next frame, and obtain a plurality of second feature points in the second face image, thereby obtaining a plurality of second feature information; Comparing the difference between the relative position information of the first human face image and the second human face image to determine the position, movement state and zoom ratio of the human face, and then calculate the positions of the plurality of second feature points; and The preset glasses models are synthesized at the positions of the plurality of second feature points. 2. The interactive service method for virtual glasses try-on according to claim 1, wherein the plurality of first feature points include the corners of the left and right eyes. 3. The interactive service method for virtual glasses try-on according to claim 2, wherein the plurality of first feature point features include both sides of the corner of the mouth. 4. The glasses virtual try-on interactive service method according to claim 1, characterized in that, according to the plurality of first feature points, it is judged whether the first human face image conforms to the logic of the human face, and if not, it is re- Obtain the first human face image or the plurality of first feature points. 5. The interactive service method for virtual glasses try-on as claimed in claim 1, wherein a search operation is applied to determine whether the plurality of first feature points are located in the frame, and if not, the first person is reacquired The face image or the plurality of first feature points. 6. The interactive service method for virtual glasses try-on according to claim 1, wherein the technical feature is pixel color information. 7. The glasses virtual try-on interactive service method according to claim 6, wherein the pixel color information is radiated in m directions from each feature point, and n pixels are obtained in each direction to obtain As color information, and the m and n are positive integers. 8. The glasses virtual try-on interactive service method according to claim 6, wherein the pixel color information is radiated in m directions by each feature point in a semicircle, and n pixels are obtained in each direction , as color information, and the m and n are positive integers and the semicircle at least covers the corner of one eye. 9. The glasses virtual try-on interactive service method according to claim 1, characterized in that, within a time interval, compare the face in the frame with the face in the next frame, and dynamically track whether the plurality of first feature points There are moving tracks. 10 . The interactive service method for virtual glasses try-on according to claim 1 , wherein a noise filtering method is applied to filter out noise in the next frame. 11 . 11. The interactive service method for three-dimensional try-on glasses according to claim 10, wherein the noise filtering method is one of Gaussian blur method, median method and mean method. 12. The glasses virtual try-on interactive service method according to claim 1, wherein a comparison range is preset to compare the plurality of first characteristic information with the plurality of second characteristic information, and take the plurality of A plurality of error values between the first characteristic information and the plurality of second characteristic information, sorting the plurality of error values, obtaining the first i minimum error values, and further obtaining the positions of the plurality of second characteristic points. 13 . The virtual interaction service method for glasses according to claim 1 , wherein, according to the positions of the plurality of first feature points and the plurality of second feature points, the inclination of the human face is calculated by slope. 14 . 14. The interactive service method for virtual glasses try-on according to claim 1, characterized in that, according to the distance between the plurality of first feature points and the distance between the plurality of second feature points, the length change of the person is calculated. The distance ratio of the face. 15. The glasses virtual try-on interactive service method according to claim 1, characterized in that, according to the ratio of the plurality of first feature points to the plurality of second feature points, the rotation angle of the human face is calculated by proportional changes and pitch angle. 16. The interactive service method for virtual glasses try-on according to claim 13, 14 or 15, wherein if the inclination, distance ratio, rotation angle or pitch angle of the human face exceeds the preset allowable value, the Obtain the second face image. 17. The glasses virtual interaction service method according to claim 1, wherein the glasses model is scaled and rotated according to the size of the face and the plurality of first feature points, and then synthesized into the face. 18. The glasses virtual interaction service method according to claim 1, wherein the glasses model is a three-dimensional glasses model. 19. The three-dimensional virtual interactive service method for glasses as claimed in claim 18, wherein a camera device is used to shoot a physical glasses from at least three directions, and obtain plane images of the three directions; and combine the three A plane image in one direction to obtain the 3D glasses model. 20. A glasses virtual try-on interactive service system, characterized in that, A capture unit locates a human face through a frame in a frame, and obtains a first human face image; The processing unit is coupled to the extraction unit and located at the feature of the eye, acquires a plurality of first feature points and feature information according to a sampling mode and stores the plurality of first feature information and the distance between the plurality of first feature points, And using dynamic image judgment and search and tracking feature information to obtain a second face image in the next frame, and obtain a plurality of second feature points in the second face image, and then obtain a plurality of second feature information; The analysis unit, coupled to the processing unit, is used to compare the difference in relative position information between the first human face image and the second human face image to determine the position, moving state and scaling ratio of the human face, and then calculate the the positions of multiple second feature points; A synthesis unit, coupled to the analysis unit, synthesizes a preset glasses model at the positions of the plurality of second feature points; and a glasses database, which stores data of the glasses model. 21. The glasses virtual try-on interactive service system according to claim 20, wherein the capture unit is a camera device. 22. The glasses virtual try-on interactive service system according to claim 20, wherein the plurality of first feature points include the corners of the left and right eyes. 23. The glasses virtual try-on interactive service system according to claim 22, wherein the plurality of first feature points also include two sides of the corners of the mouth. 24. The glasses virtual try-on interactive service system according to claim 20, wherein the processing unit features include judging whether the first face image conforms to the face logic based on a plurality of first feature points, and if not If it matches, the first human face image or the plurality of first feature points are reacquired. 25. The glasses virtual try-on interactive service system according to claim 20, wherein the processing unit includes applying a search operation to determine whether the plurality of first feature points are located within the frame, and if not, then Reacquiring the first face image or the plurality of first feature points. 26. The glasses virtual try-on interactive service system according to claim 20, wherein the feature information is pixel color information, and the processing unit radiates m directions from each feature point, and in each Obtain n pixels in the direction as color information, wherein m and n are positive integers. 27. The glasses virtual try-on interactive service system according to claim 20, wherein the feature information is pixel color information, and the processing unit radiates m directions in a semicircle from each of the feature points, and each n pixels are obtained in one direction as color information, and the semicircle covers at least one eye corner, wherein m and n are positive integers. 28. The glasses virtual try-on interactive service system according to claim 20, wherein the analysis unit compares the face in the frame with the face in the next frame in a time interval, and tracks the plurality of first features Whether the point has a moving track, and apply a noise filtering method to filter out the noise in the next frame. 29. The interactive service system for virtual glasses try-on according to claim 28, wherein the noise filtering method is one of Gaussian blur method, median method and mean method. 30. The glasses virtual try-on interactive service system according to claim 20, wherein the analysis unit presets a comparison range to compare the plurality of first characteristic information with the plurality of second characteristic information, and take A plurality of error values between the plurality of first feature information and the plurality of second feature information, and sorting the plurality of error values, and obtaining the first i minimum error values, and then obtaining the positions of the plurality of second feature points . 31. The glasses virtual try-on interactive service system according to claim 20, characterized in that, the analysis unit calculates the slope of the face according to the positions of the plurality of first feature points and the plurality of second feature points. Inclination, according to the distance between the plurality of first feature points and the distance between the plurality of second feature points, calculate the distance ratio of the face according to the length change and according to the plurality of first feature points and the plurality of second feature points The ratio of the two feature points is used to calculate the rotation angle and pitch angle of the face, and if the inclination, distance ratio, rotation angle or pitch angle of the face exceeds the preset allowable value, the second person is reacquired face image. 32. The glasses virtual try-on interactive service system according to claim 20, wherein the synthesis unit scales and rotates the glasses model according to the size of the human face and the plurality of first feature points, and then synthesizes the glasses model into The face. 33. A virtual glasses try-on interactive service method, characterized in that, Locating a human face through a frame in a frame, and obtaining a first human face image; At the feature of the pupil of the eye, a plurality of first feature points and feature information are obtained according to a sampling mode, and the distance between the plurality of first feature information and the plurality of first feature points is stored; Obtaining the second face image in the next frame by using dynamic image judgment and searching and tracking feature information, and obtaining a plurality of second feature points in the second face image, and then obtaining a plurality of second feature information; Comparing the difference between the relative position information of the first human face image and the second human face image to determine the position, movement state and zoom ratio of the human face, and then calculate the positions of the plurality of second feature points; and A preset contact lens model is synthesized at the positions of the plurality of second feature points. 34. The interactive service method for virtual glasses try-on according to claim 33, wherein the plurality of first feature point features include both sides of the corner of the mouth. 35. The glasses virtual try-on interactive service method according to claim 33, characterized in that, according to a plurality of first feature points, it is judged whether the first human face image conforms to the human face logic, and if not, it is re-obtained The first face image or the plurality of first feature points. 36. The interactive service method for virtual glasses try-on according to claim 33, wherein a search operation is applied to determine whether the plurality of first feature points are located within the frame, and if not, the first feature points are reacquired. A face image or the plurality of first feature points. 37. The interactive service method for virtual glasses try-on as claimed in claim 33, wherein the characteristic information is pixel color information, and the pixel color information is radiated in m directions from each of the characteristic points, and each Obtain n pixels in one direction as color information, and m and n are positive integers. 38. The glasses virtual try-on interactive service method according to claim 33, wherein, within a time interval, compare the face in the frame with the face in the next frame, and dynamically track whether the plurality of first feature points There is a moving track, and a noise filtering method is applied to filter out the noise in the next frame. 39. The glasses virtual try-on interactive service method according to claim 33, wherein a comparison range is preset to compare the plurality of first characteristic information with the plurality of second characteristic information, and take the plurality of A plurality of error values between the first characteristic information and the plurality of second characteristic information, sorting the plurality of error values, obtaining the first i minimum error values, and further obtaining the positions of the plurality of second characteristic points. 40. The glasses virtual try-on interactive service method according to claim 33, characterized in that, according to the positions of the plurality of first feature points and the plurality of second feature points, the inclination of the human face is calculated by slope, According to the distance between the plurality of first feature points and the distance between the plurality of second feature points, the distance ratio of the human face is calculated according to the length change and according to the plurality of first feature points and the plurality of second feature points Calculate the rotation angle and pitch angle of the face according to the ratio, and if the face tilt, distance ratio, rotation angle or pitch angle exceeds a preset allowable value, the second face image is reacquired. 41. The glasses virtual try-on interactive service method according to claim 33, wherein the glasses model is scaled and rotated according to the size of the face and the plurality of first feature points, and then synthesized into the face .

Images