Mode judgment system based on human face recognition

Abstract

The invention relates to a mode judgment system based on human face recognition, and the system comprises a content capturing device which is used for carrying out the screen capturing of a live broadcast image of a set live broadcast room into which a portable terminal enters at present; and the mode judgment mechanism is used for judging that the set direct broadcasting room is suspected to start a buffing and beautifying mode to trigger subsequent manual secondary confirmation when the entity smoothness level corresponding to the anchor face determined based on the intelligent face recognition result exceeds the limit. The mode judgment system based on human face recognition is intelligent in operation and wide in application. The entity smoothness level corresponding to the face of the anchor can be intelligently identified based on various visual data, and when the determined entity smoothness level corresponding to the face of the anchor is greater than or equal to the set level threshold value, it is judged that the set live broadcast room where the anchor is located is suspected to start the buffing beautifying mode so as to trigger subsequent manual secondary confirmation. Therefore, intelligent judgment on whether the buffing and beautifying mode is started on the live broadcast screen capture picture is completed.

Description

Technical Field

[0001] This invention relates to the field of facial recognition, and more particularly to a pattern recognition system based on human facial recognition. Background Technology

[0002] Like other biometric features (fingerprints, irises, etc.), the human face is innate. Its uniqueness and resistance to replication provide the necessary prerequisites for identity verification. Compared with other types of biometrics, facial recognition has the following characteristics: Non-coercive: Users do not need to cooperate with the facial acquisition device; facial images can be obtained almost unconsciously, and this sampling method is not "coercive"; Non-contact: Users do not need to directly contact the device to obtain facial images; Concurrency: In practical application scenarios, multiple faces can be sorted, judged, and recognized; In addition, it also conforms to visual characteristics: the characteristic of "recognizing people by their appearance," as well as the characteristics of simple operation, intuitive results, and good concealment.

[0003] Viewers watching live streams of hosts showcasing their talents in different studios are often confused about whether the hosts are displaying their real faces or using skin-smoothing and beautifying filters. Current technologies either rely on visual inspection of the host's face to determine if it's too smooth, or use simpler methods to identify whether a host is using skin-smoothing or beautifying filters, but the results are often unsatisfactory. Summary of the Invention

[0004] To address the aforementioned technical problems, this invention provides a pattern recognition system based on human face recognition, the system comprising:

[0005] A content capture device, installed in a portable terminal, is used to capture the live screen of the live room currently accessed by the portable terminal, so as to obtain and output the corresponding real-time screenshot.

[0006] A gamma correction device, installed in a portable terminal and connected to the content capture device, is used to perform gamma correction processing on the received real-time screenshot to obtain and output the corresponding gamma-corrected image.

[0007] An interference cancellation device, connected to the gamma correction device, is used to perform pulse interference cancellation processing on the received gamma correction image to obtain and output the corresponding interference cancellation image.

[0008] A signal filtering device, connected to the interference cancellation device, is used to perform maximum value filtering on the received interference cancellation image to obtain and output the corresponding filtered image.

[0009] A directional analysis mechanism, installed within a portable terminal and connected to the signal filtering device, is used to uniformly divide the image region corresponding to the anchor's face in the received filtered image to obtain image sub-regions of equal size corresponding to the anchor's face. It calculates the overall depth of field of each image sub-region corresponding to the anchor's face and determines the entity smoothness level of the anchor's face based on the numerical difference data of the overall depth of field corresponding to each image sub-region. The smaller the numerical difference data, the higher the determined entity smoothness level of the anchor's face. The overall depth of field of each image sub-region corresponding to the anchor's face is the arithmetic mean of the depth values ​​corresponding to each pixel within the image sub-region. Furthermore, the higher the maximum noise amplitude of the image region corresponding to the anchor's face in the filtered image, the fewer the number of image sub-regions corresponding to the anchor's face obtained by uniformly dividing the image region corresponding to the anchor's face in the filtered image.

[0010] The mode determination mechanism, connected to the orientation analysis mechanism, is used to determine that the set live broadcast room is suspected of having a skin smoothing and beautification mode when the physical smoothness level corresponding to the determined anchor's face is greater than or equal to a set level threshold, thereby triggering subsequent manual secondary confirmation. It is also used to determine that the set live broadcast room is not having a skin smoothing and beautification mode when the physical smoothness level corresponding to the determined anchor's face is less than the set level threshold.

[0011] The pattern judgment system based on human face recognition of the present invention is intelligent in operation and widely applicable. It can intelligently identify the smoothness level of the entity corresponding to the streamer's face based on various visual data, and when the determined smoothness level of the entity corresponding to the streamer's face is greater than or equal to a set level threshold, it determines that the streamer's live broadcast room is suspected of having a skin-smoothing beautification mode enabled, triggering subsequent manual secondary confirmation. This completes the intelligent judgment of whether a skin-smoothing beautification mode is enabled in the live broadcast screenshot. Attached Figure Description

[0012] The embodiments of the present invention will now be described with reference to the accompanying drawings.

[0013] Figure 1 This is a schematic diagram of the internal structure of a pattern judgment system based on human face recognition according to the first embodiment of the present invention.

[0014] Figure 2 This is a schematic diagram of the internal structure of a pattern judgment system based on human face recognition according to a second embodiment of the present invention.

[0015] Figure 3 This is a schematic diagram of the internal structure of a pattern judgment system based on human face recognition according to a third embodiment of the present invention. Detailed Implementation

[0016] The embodiments of the pattern judgment system based on human face recognition of the present invention will now be described in detail with reference to the accompanying drawings.

[0017] First Embodiment

[0018] Figure 1 This is a schematic diagram of the internal structure of a pattern determination system based on human face recognition according to a first embodiment of the present invention. The system includes:

[0019] A content capture device, installed in a portable terminal, is used to capture the live screen of the live room currently accessed by the portable terminal, so as to obtain and output the corresponding real-time screenshot.

[0020] Specifically, the content capture device is installed in the portable terminal and is used to capture the live screen of the set live room currently entered by the portable terminal to obtain and output the corresponding real-time screenshot. The content capture device completes the capture of the live screen of the set live room currently entered by the portable terminal from the cache chip of the portable terminal.

[0021] A gamma correction device, installed in a portable terminal and connected to the content capture device, is used to perform gamma correction processing on the received real-time screenshot to obtain and output the corresponding gamma-corrected image.

[0022] An interference cancellation device, connected to the gamma correction device, is used to perform pulse interference cancellation processing on the received gamma correction image to obtain and output the corresponding interference cancellation image.

[0023] A signal filtering device, connected to the interference cancellation device, is used to perform maximum value filtering on the received interference cancellation image to obtain and output the corresponding filtered image.

[0024] A directional analysis mechanism, installed within a portable terminal and connected to the signal filtering device, is used to uniformly divide the image region corresponding to the anchor's face in the received filtered image to obtain image sub-regions of equal size corresponding to the anchor's face. It calculates the overall depth of field of each image sub-region corresponding to the anchor's face and determines the entity smoothness level of the anchor's face based on the numerical difference data of the overall depth of field corresponding to each image sub-region. The smaller the numerical difference data, the higher the determined entity smoothness level of the anchor's face. The overall depth of field of each image sub-region corresponding to the anchor's face is the arithmetic mean of the depth values ​​corresponding to each pixel within the image sub-region. Furthermore, the higher the maximum noise amplitude of the image region corresponding to the anchor's face in the filtered image, the fewer the number of image sub-regions corresponding to the anchor's face obtained by uniformly dividing the image region corresponding to the anchor's face in the filtered image.

[0025] The mode judgment mechanism, connected to the orientation analysis mechanism, is used to determine that the set live room is suspected of having a skin smoothing and beautification mode when the physical smoothness level of the determined anchor's face is greater than or equal to a set level threshold, so as to trigger subsequent manual secondary confirmation. It is also used to determine that the set live room has not had a skin smoothing and beautification mode when the physical smoothness level of the determined anchor's face is less than the set level threshold.

[0026] The process involves uniformly dividing the image region corresponding to the anchor's face in the received filtered image to obtain image sub-regions of equal size, calculating the overall depth of field of each image sub-region corresponding to the anchor's face, and determining the entity smoothness level of the anchor's face based on the numerical difference data of the overall depth of field of each image sub-region corresponding to the anchor's face. This includes identifying the image region corresponding to the anchor's face in the received filtered image based on the outline features of the anchor's face.

[0027] Second Embodiment

[0028] Figure 2 This is a schematic diagram of the internal structure of a pattern judgment system based on human face recognition according to a second embodiment of the present invention.

[0029] Compared to Figure 1 The pattern determination system based on human face recognition shown in the second embodiment of the present invention may further include:

[0030] A current sensing device is disposed near the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism, and is respectively connected to the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism;

[0031] Specifically, the current sensing device, which is located near the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism and is respectively connected to the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism, includes: using a GAL device to implement the current sensing device, and being located near the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism and respectively connected to the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism;

[0032] The current sensing device, which is located near the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism and is connected to the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism respectively, includes: the current sensing device being used to perform on-site measurement of the current current of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism respectively.

[0033] Third Embodiment

[0034] Figure 3 This is a schematic diagram of the internal structure of a pattern judgment system based on human face recognition according to a third embodiment of the present invention.

[0035] Compared to Figure 1 The pattern recognition system based on human face recognition shown in the third embodiment of the present invention may further include:

[0036] A humidity detection device is disposed near the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism, and is respectively connected to the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism;

[0037] The humidity detection device, which is located near the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism and is connected to the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism respectively, includes: the humidity detection device being used to perform on-site measurement of the current humidity of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism respectively.

[0038] Next, the specific structure of the pattern judgment system based on human face recognition of the present invention will be further described.

[0039] In the various pattern determination systems based on human face recognition according to embodiments of the present invention:

[0040] An FPGA is used to perform image data processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism.

[0041] In the various pattern determination systems based on human face recognition according to embodiments of the present invention:

[0042] Using an FPGA to perform image data processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism includes: performing directional filtering processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism.

[0043] In the various pattern determination systems based on human face recognition according to embodiments of the present invention:

[0044] Using an FPGA to perform image data processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism includes: performing box filtering processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism.

[0045] In the various pattern determination systems based on human face recognition according to embodiments of the present invention:

[0046] Using an FPGA to perform image data processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism includes: performing wavelet filtering processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism.

[0047] And in the various pattern determination systems based on human face recognition according to embodiments of the present invention:

[0048] Using an FPGA to perform image data processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism includes: performing bilinear interpolation processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism.

[0049] Furthermore, in the aforementioned pattern judgment system based on human face recognition, the image region corresponding to the anchor's face in the received filtered image is uniformly divided into image sub-regions of equal size corresponding to the anchor's face. The overall depth of field of each image sub-region corresponding to the anchor's face is calculated. The determination of the entity smoothness level corresponding to the anchor's face based on the numerical difference data of the overall depth of field corresponding to each image sub-region corresponding to the anchor's face also includes: using a numerical conversion function to represent the numerical conversion relationship between the determined entity smoothness level corresponding to the anchor's face and the numerical difference data of the overall depth of field corresponding to each image sub-region corresponding to the anchor's face.

[0050] Therefore, it can be seen that the present invention has the following significant technical effects:

[0051] First, the image region corresponding to the anchor's face in the received filtered image is uniformly divided into blocks to obtain image sub-regions of equal size corresponding to the anchor's face. The higher the maximum noise amplitude of the image region corresponding to the anchor's face in the filtered image, the fewer the number of image sub-regions corresponding to the anchor's face obtained by uniformly dividing the image region corresponding to the anchor's face in the filtered image.

[0052] Secondly: Calculate the overall depth of field of each image sub-region corresponding to the anchor's face, and determine the entity smoothing level of the anchor's face based on the numerical difference data of the overall depth of field of each image sub-region corresponding to the anchor's face. The smaller the numerical difference data, the higher the entity smoothing level of the anchor's face. The overall depth of field of each image sub-region corresponding to the anchor's face is the arithmetic mean of the depth values ​​of each pixel in the image sub-region.

[0053] Furthermore, the reference pattern judgment mechanism is used to determine if the smoothness level of the entity corresponding to the anchor's face is greater than or equal to a set level threshold, thereby triggering subsequent manual secondary confirmation. It is also used to determine if the smoothness level of the entity corresponding to the anchor's face is less than the set level threshold, thereby realizing intelligent judgment of the live broadcast mode based on the recognition of the smoothness level of human face in the live broadcast screenshot.

[0054] Those skilled in the art will understand that various modifications and alterations can be made without departing from the scope and spirit of this invention. Therefore, it should be understood that the above embodiments are for illustrative purposes only and are not intended to limit the scope. Because the scope of this invention is defined by the claims rather than the foregoing description, any changes and modifications falling within the scope and boundaries of the claims, or their equivalents, are subject to the claims.

Claims

1. A pattern recognition system based on human face recognition, characterized in that, The system includes: A content capture device, installed in a portable terminal, is used to capture the live screen of the live room currently accessed by the portable terminal, so as to obtain and output the corresponding real-time screenshot. A gamma correction device, installed in a portable terminal and connected to the content capture device, is used to perform gamma correction processing on the received real-time screenshot to obtain and output the corresponding gamma-corrected image. An interference cancellation device, connected to the gamma correction device, is used to perform pulse interference cancellation processing on the received gamma correction image to obtain and output the corresponding interference cancellation image. A signal filtering device, connected to the interference cancellation device, is used to perform maximum value filtering on the received interference cancellation image to obtain and output the corresponding filtered image. A directional analysis mechanism, installed within a portable terminal and connected to the signal filtering device, is used to uniformly divide the image region corresponding to the anchor's face in the received filtered image to obtain image sub-regions of equal size corresponding to the anchor's face. It calculates the overall depth of field of each image sub-region corresponding to the anchor's face and determines the entity smoothness level of the anchor's face based on the numerical difference data of the overall depth of field corresponding to each image sub-region. The smaller the numerical difference data, the higher the determined entity smoothness level of the anchor's face. The overall depth of field of each image sub-region corresponding to the anchor's face is the arithmetic mean of the depth values ​​corresponding to each pixel within the image sub-region. Furthermore, the higher the maximum noise amplitude of the image region corresponding to the anchor's face in the filtered image, the fewer the number of image sub-regions corresponding to the anchor's face obtained by uniformly dividing the image region corresponding to the anchor's face in the filtered image. The mode determination mechanism, connected to the orientation analysis mechanism, is used to determine that the set live broadcast room is suspected of having a skin smoothing and beautification mode when the physical smoothness level corresponding to the determined anchor's face is greater than or equal to a set level threshold, thereby triggering subsequent manual secondary confirmation. It is also used to determine that the set live broadcast room is not having a skin smoothing and beautification mode when the physical smoothness level corresponding to the determined anchor's face is less than the set level threshold.

2. The pattern recognition system based on human face recognition as described in claim 1, characterized in that: The image region corresponding to the anchor's face in the received filtered image is uniformly divided into blocks to obtain image sub-regions of equal size corresponding to the anchor's face. The overall depth of field of each image sub-region corresponding to the anchor's face is calculated. The entity smoothness level corresponding to the anchor's face is determined based on the numerical difference data of the overall depth of field corresponding to each image sub-region corresponding to the anchor's face. This includes: identifying the image region corresponding to the anchor's face in the received filtered image based on the outline features of the anchor's face.

3. The pattern recognition system based on human face recognition as described in claim 2, characterized in that, The system also includes: A current sensing device is disposed near the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism, and is respectively connected to the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism; The current sensing device, which is located near the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism and is connected to the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism respectively, includes: the current sensing device being used to perform on-site measurement of the current current of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism respectively.

4. The pattern recognition system based on human face recognition as described in claim 2, characterized in that, The system also includes: A humidity detection device is disposed near the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism, and is respectively connected to the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism; The humidity detection device, which is located near the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism and is connected to the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism respectively, includes: the humidity detection device being used to perform on-site measurement of the current humidity of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism respectively.

5. The pattern recognition system based on human face recognition as described in any one of claims 2-4, characterized in that: An FPGA is used to perform image data processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism.

6. The pattern recognition system based on human face recognition as described in claim 5, characterized in that: Using an FPGA to perform image data processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism includes: performing directional filtering processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism.

7. The pattern recognition system based on human face recognition as described in claim 5, characterized in that: Using an FPGA to perform image data processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism includes: performing box filtering processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism.

8. The pattern recognition system based on human face recognition as described in claim 5, characterized in that: Using an FPGA to perform image data processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism includes: performing wavelet filtering processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism.

9. The pattern recognition system based on human face recognition as described in claim 5, characterized in that: Using an FPGA to perform image data processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism includes: performing bilinear interpolation processing on the output data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism to obtain the corresponding output processing data of the gamma correction device, the interference cancellation device, the signal filtering device, and the directional analysis mechanism.

Images