End-side cooperative face recognition method and system, computer device and medium

Abstract

The application discloses a kind of based on end side collaborative face recognition method, system, computer equipment and storage medium, its method implementation, comprising: by first terminal equipment to the user to be identified face recognition;When first terminal face recognition fails, by first terminal to dispatch module sends the face image data of user to be identified, face image data includes face feature;Face feature is compared with all registered face features, to obtain N registered face feature codes that meet preset matching condition;Determine the second terminal to which the registered face feature code belongs, and send face image data to the second terminal and / or server, for feature comparison;Receive the comparison result sent by the second terminal and / or server, and send to first terminal, so that the first terminal outputs final face recognition result according to the comparison result.Can support large-capacity face recognition, effectively solve the problem of insufficient computing power and insufficient running memory of single terminal.

Description

Technical Field

[0001] This invention relates to the field of facial recognition technology, and in particular to a method, system, computer device, and storage medium for edge-side collaborative facial recognition. Background Technology

[0002] Facial recognition, as a non-contact biometric technology, is one of the most commonly used modalities in the field of biometric identification. It is characterized by its visualization, convenience, and alignment with human thinking habits, and has been widely used in the field of public safety in recent years.

[0003] Existing facial recognition systems all adopt a standalone deployment model. The insufficient computing power and memory of a single device prevent the deployment of large-capacity facial recognition algorithms. Furthermore, because a single device cannot store enough registered facial data, it cannot recognize a large number of faces. With the widespread adoption and rapid development of facial recognition, how to quickly identify the ever-increasing volume of facial data has become a pressing issue. Summary of the Invention

[0004] Therefore, it is necessary to provide a face recognition method, system, computer device, and storage medium based on edge-side collaborative technology to address the above-mentioned technical problems. This would solve the problem that in the existing technology, the insufficient computing power and running memory of a single device prevent the deployment of large-capacity face recognition algorithms and the storage of more registered face data, thus hindering the recognition of more faces.

[0005] Firstly, a face recognition method based on edge-side collaborative technology is provided, including:

[0006] The first terminal device performs facial recognition on the user to be identified;

[0007] When the face recognition of the first terminal fails, the face image data of the user to be identified is sent to the scheduling module through the first terminal. The face image data includes face features.

[0008] The facial features are compared with all registered facial features to obtain N registered facial feature codes that meet preset matching conditions;

[0009] The second terminal to which the registered facial feature code belongs is determined, and the facial image data is sent to the second terminal and / or the server for feature comparison.

[0010] The system receives the comparison results sent by the second terminal and / or the server, and sends them to the first terminal so that the first terminal outputs the final face recognition result based on the comparison results.

[0011] In one embodiment, before performing facial recognition on the user to be identified via the first terminal device, the process includes:

[0012] According to preset rules, different registered facial features are distributed to multiple pre-deployed terminals so that the terminals can perform facial recognition using the registered facial features.

[0013] In one embodiment, comparing the facial feature with all registered facial features to obtain N registered facial feature codes that meet preset matching conditions includes:

[0014] Calculate the comprehensive matching score between the stated facial feature and all registered facial features;

[0015] The comprehensive matching score is compared with a preset score threshold to obtain registered face features whose comprehensive matching score is greater than the preset score threshold.

[0016] Among the registered face features whose matching scores are greater than the preset score threshold, N registered face feature codes are obtained in descending order of the comprehensive matching scores.

[0017] In one embodiment, receiving the comparison result sent by the second terminal and / or the server and sending it to the first terminal includes:

[0018] After receiving the comparison results sent by the second terminal and the server, it is determined whether the comparison results are consistent;

[0019] If the comparison results are inconsistent, the facial features are compared again with all registered facial features to obtain updated registered facial feature codes.

[0020] Based on the updated registered facial feature code, a third terminal is identified, and the facial feature data and the registered facial feature code are sent to the third terminal.

[0021] In one embodiment, the second terminal includes multiple terminals, and before receiving the comparison result sent by the second terminal and / or the server and sending it to the first terminal, the process includes...

[0022] After receiving multiple comparison results sent by the second terminal, determine whether there is a successful comparison result among the comparison results;

[0023] If the judgment result is yes, then send the successful comparison result and the successfully registered face feature code to the first terminal;

[0024] If the judgment result is negative, a comparison failure result is sent to the first terminal.

[0025] In one embodiment, the step of performing facial recognition on the user to be identified via the first terminal device includes:

[0026] The facial image data of the user to be identified is obtained through the first terminal;

[0027] Feature extraction is performed on the facial image data to obtain the facial features of the user to be identified;

[0028] The facial features are compared with the registered facial features corresponding to the first terminal;

[0029] When the matching degree between the registered facial features and the facial features reaches a preset threshold, the recognition is successful; otherwise, the recognition fails.

[0030] In one embodiment, the step of extracting features from the facial image data to obtain the facial features of the user to be identified includes:

[0031] Perform face detection on the face image data;

[0032] If a face is detected, facial landmark detection is performed on the face image data;

[0033] Based on the results of facial landmark detection, the facial image data is corrected, and the corrected facial image data is adjusted to a facial image of a preset size.

[0034] Image enhancement is performed on the face image of the preset size;

[0035] The enhanced face image is input into the feature model to obtain the face features.

[0036] Secondly, an edge-side collaborative face recognition system is provided, including:

[0037] A face recognition unit is used to perform face recognition on the user to be identified through a first terminal device;

[0038] A face image data acquisition unit is used to send the face image data of the user to be identified to the scheduling module through the first terminal when the face recognition of the first terminal fails. The face image data includes face features.

[0039] The registered face feature code acquisition unit is used to compare the face feature with all registered face features to obtain N registered face feature codes that meet preset matching conditions.

[0040] The feature comparison unit is used to determine the second terminal to which the registered face feature code belongs, and to send the face image data to the second terminal and / or the server for feature comparison.

[0041] A face recognition result output unit is used to receive the comparison result sent by the second terminal and / or the server, and send it to the first terminal so that the first terminal outputs the final face recognition result based on the comparison result.

[0042] Thirdly, a computer device is provided, including a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, wherein the processor, when executing the computer-readable instructions, implements the steps of the edge-based collaborative face recognition method described above.

[0043] Fourthly, one or more readable storage media are provided, the readable storage media storing computer-readable instructions, which, when executed by a processor, implement the steps of the edge-based collaborative face recognition method described above.

[0044] The aforementioned end-to-end collaborative face recognition method, system, computer device, and storage medium are implemented as follows: The method includes: performing face recognition on a user to be identified via a first terminal device; when the first terminal fails to recognize the user's face, sending the face image data of the user to be identified, including face features, to a scheduling module via the first terminal device; comparing the face features with all registered face features to obtain N registered face feature codes that meet preset matching conditions; determining the second terminal to which the registered face feature codes belong, and sending the face image data to the second terminal and / or a server for feature comparison; receiving the comparison results sent by the second terminal and / or the server, and sending them to the first terminal so that the first terminal outputs the final face recognition result based on the comparison results. In this application, due to the limited storage capacity and computing power of the terminal device, only a portion of the registered facial feature codes are stored. When the first terminal fails to recognize the face, the facial feature data of the user to be recognized can be sent to the scheduling module for comparison with all registered facial features. N registered facial feature codes that meet the preset comparison conditions are obtained and sent to the corresponding terminal and / or server for comparison. The comparison result is then fed back to the first terminal as the final facial recognition result. This eliminates the need to store all registered facial features in each terminal device and enables collaborative facial recognition between different devices. It supports large-capacity facial recognition and effectively solves the problems of insufficient computing power and insufficient running memory. Attached Figure Description

[0045] To more clearly illustrate the technical solutions of the embodiments of the present invention, the drawings used in the description of the embodiments of the present invention will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0046] Figure 1 This is a schematic diagram of an application environment for an end-to-end collaborative face recognition method according to an embodiment of the present invention;

[0047] Figure 2 This is a flowchart illustrating an embodiment of the edge-side collaborative face recognition method of the present invention. Figure 1 ;

[0048] Figure 3 This is a schematic flowchart of an end-to-end collaborative face recognition method according to an embodiment of the present invention. Figure 2 ;

[0049] Figure 4 This is a flowchart illustrating a facial feature extraction method according to an embodiment of the present invention;

[0050] Figure 5 This is a schematic diagram of a terminal-side collaborative face recognition system according to an embodiment of the present invention;

[0051] Figure 6 This is a schematic diagram of a computer device according to an embodiment of the present invention. Detailed Implementation

[0052] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of the present invention. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0053] The edge-side collaborative face recognition method provided in this embodiment can be applied to, for example... Figure 1 In this application environment, the terminal can communicate with the scheduling module, which in turn can communicate with the server.

[0054] The terminal can be a terminal device deployed within a unified local area network. This device is not limited to specific hardware identification devices, such as personal computers, laptops, smartphones, tablets, and portable wearable devices. It can also refer to the deployed server.

[0055] The server side can be implemented using a standalone server or a server cluster consisting of multiple servers.

[0056] In one embodiment, such as Figure 2 As shown, an edge-side collaborative face recognition method is provided, including the following steps:

[0057] In step S110, facial recognition of the user to be identified is performed through the first terminal device;

[0058] In this embodiment, the first terminal device is equipped with a camera. The camera captures video data of the user being identified within a target acquisition area during face recognition, and extracts image frames from the video data. The image containing the user's face is used as the image to be identified. Feature extraction is performed on the image to be identified, allowing comparison with pre-stored registered facial features in the first terminal device. If the comparison is successful, recognition is successful; otherwise, recognition fails.

[0059] The camera device is either a visible camera or an infrared camera.

[0060] In this embodiment of the application, each terminal device may store different registered facial features.

[0061] In step S120, when face recognition fails, the face image data of the user to be identified is sent to the scheduling module through the first terminal. The face image data includes face features.

[0062] In this embodiment of the application, when the first terminal fails to recognize the face of the user to be recognized, the first terminal will send the facial feature data of the user to be recognized that it has collected to the scheduling module, such as face images, extracted facial features, and other data.

[0063] In step S130, the facial feature is compared with all registered facial features to obtain N registered facial feature codes that meet preset matching conditions;

[0064] In this embodiment, the scheduling module can be a server or a terminal device. The scheduling module can communicate with a database that stores all registered facial features. These registered facial features are the facial feature data of users who have already registered their identities.

[0065] In the embodiments of this application, N is a natural number, such as 1, 10, 50, 100, etc., which can be set according to the actual situation.

[0066] In this embodiment of the application, the registered facial feature code can be used to identify the same user, that is, the facial feature code of the same registered user is the same. Specifically, the registered facial feature code can be encoded by the identity registration platform according to the preset encoding rules after the user registers his or her identity, such as 0001, 0002, etc.

[0067] In this embodiment of the application, after the scheduling module obtains the facial features of the user to be identified sent by the first terminal, it can compare the facial features with all registered facial features, calculate the matching value between the facial features and each registered facial feature, and when the matching value meets the preset matching threshold, it can obtain N registered facial features that meet the preset matching threshold and their corresponding codes.

[0068] In step S140, the second terminal to which the registered face feature code belongs is determined, and the face image data is sent to the second terminal and / or the server for feature comparison.

[0069] In this embodiment, each registered user's facial features have a unique code to identify the user, and each terminal can store different facial features. That is, the codes of registered facial features stored in each terminal are different. Therefore, after obtaining N registered facial feature codes, they can be assigned to the corresponding second terminal according to the registered facial feature codes. For example, the registered facial feature codes stored in terminal A are 1-1000, those stored in terminal B are 1001-2000, and those stored in terminal C are 2001-3000. N is 2, and the obtained registered facial feature codes are 20 and 2380. Then, terminal A and terminal C are used as second terminals, and the facial image data of the user to be identified and the registered facial feature codes are sent to terminal A and terminal C to perform feature comparison through terminal A and terminal C.

[0070] The second terminal can be one or more, depending on the location where the registered facial feature codes are stored.

[0071] In this embodiment of the application, in addition to sending the facial image features of the user to be identified to the corresponding second terminal for comparison, the facial image features of the user to be identified can also be sent to the server. The server can store all the facial features to be registered, and feature comparison can also be achieved through the server.

[0072] Specifically, the target terminal for transmission can be determined based on the busy status of both the second terminal and the server. For example, when the second terminal is busy, the facial image features of the user to be identified are sent to the server first. When the server is busy, the facial image features of the user to be identified are sent to the second terminal first.

[0073] In step S150, the comparison result sent by the second terminal and / or the server is received and sent to the first terminal so that the first terminal outputs the final face recognition result based on the comparison result.

[0074] In this embodiment of the application, the comparison result may include a successful comparison or a failed comparison. When the comparison is successful, the face feature code of the successful comparison can be sent to the first terminal and the face recognition is successful. When the comparison fails, the result of the failed comparison can be sent to the first terminal and the face recognition fails is output.

[0075] This application provides an edge-side collaborative face recognition method, comprising: performing face recognition on a user to be recognized through a first terminal device; when the face recognition fails on the first terminal, sending the face image data of the user to be recognized, the face image data including face features, to a scheduling module through the first terminal device; comparing the face features with all registered face features to obtain N registered face feature codes that meet preset matching conditions; determining the second terminal to which the registered face feature codes belong, and sending the face image data to the second terminal and / or a server for feature comparison; receiving the comparison results sent by the second terminal and / or the server, and sending them to the first terminal so that the first terminal outputs the final face recognition result based on the comparison results. In this application, due to the limited storage capacity and computing power of the terminal device, only a portion of the registered facial feature codes are stored. When the first terminal fails to recognize the face, the facial feature data of the user to be recognized can be sent to the scheduling module for comparison with all registered facial features. N registered facial feature codes that meet the preset comparison conditions are obtained and sent to the corresponding terminal and / or server for comparison. The comparison result is then fed back to the first terminal as the final facial recognition result. This eliminates the need to store all registered facial features in each terminal device and enables collaborative facial recognition between different devices. It supports large-capacity facial recognition and effectively solves the problems of insufficient computing power and insufficient running memory.

[0076] In one embodiment, an edge-based collaborative face recognition method is provided, comprising:

[0077] In step S110, facial recognition of the user to be identified is performed through the first terminal device;

[0078] In this embodiment of the application, before performing facial recognition on the user to be identified through the first terminal device, the following steps are included:

[0079] According to preset rules, different registered facial features are distributed to multiple pre-deployed terminals so that the terminals can perform facial recognition using the registered facial features.

[0080] Specifically, for a high-capacity facial recognition system, before facial recognition is performed, the registered facial image features can be distributed to the terminal devices. For example, if 100,000 faces have been registered, and the facial information of these 100,000 people can be stored in the scheduling module, each device has a registration limit and processing capacity of 10,000 people. Ten devices can be deployed within the local area network to distribute different registered facial image features. For example, device 0 distributes facial IDs from 1 to 10,000, device 1 distributes facial IDs from 10,001 to 20,000, and so on. This allows each terminal device to perform facial recognition within its own computing power, solving the problems of insufficient computing power and insufficient running memory. Furthermore, through the scheduling module, the capabilities of all terminal devices can be centralized in parallel, giving full play to the segmented processing capabilities between devices.

[0081] See Figure 3 The step of performing facial recognition on the user to be identified through the first terminal device includes:

[0082] In step S210, the facial image data of the user to be identified is acquired through the first terminal;

[0083] In step S220, feature extraction is performed on the face image data to obtain the face features of the user to be identified;

[0084] In step S230, the facial features are compared with the registered facial features corresponding to the first terminal;

[0085] In step S240, if the matching degree between the registered facial feature and the facial feature reaches a preset threshold, the recognition is successful; otherwise, the recognition fails.

[0086] In this embodiment, the first terminal device is equipped with a camera. The camera captures video data of the user being identified within a target acquisition area during face recognition, and extracts image frames from the video data, using the image containing the user's face as the image to be identified. Facial landmark detection is performed on the image to be identified, and corresponding corrections are made based on the detection results. The corrected image is then input into a preset feature model for feature extraction, and compared with pre-stored registered facial features in the first terminal device. If the comparison is successful, recognition is successful; otherwise, recognition fails.

[0087] The camera device is either a visible camera or an infrared camera.

[0088] See Figure 4 In this embodiment of the application, the step of extracting features from the facial image data to obtain the facial features of the user to be identified includes:

[0089] In step S310, face detection is performed on the face image data;

[0090] If a face is detected, facial landmark detection is performed on the face image data;

[0091] In step S320, the face image data is corrected based on the results of the face key point detection, and the corrected face image data is adjusted to a face image of a preset size.

[0092] In step S340, image enhancement is performed on the face image of the preset size;

[0093] In step S330, the enhanced face image is input into the feature model to obtain the face features.

[0094] In this embodiment, after acquiring the face image of the user to be identified through the first terminal, face detection can be performed by training a network model through deep learning or by training a model through other traditional methods. When a face is detected in the face image, key points can be detected on the detected face, such as face angle, position of various parts of the face, symmetry, etc. Then, the face image can be corrected according to the detection results, for example, adjusting the tilt of the face. After the correction is completed, the corrected face image can be further normalized to a preset size, such as 128x128. Then, image enhancement operations can be performed on the face image of the preset size to improve face recognition. After image enhancement, the face image is input into a pre-trained feature model for feature extraction of a preset dimension, such as 512 dimensions. Finally, the extracted features can be stored for comparison with the pre-stored registered face features in the first terminal device.

[0095] In step S120, when face recognition fails, the face image data of the user to be identified is sent to the scheduling module through the first terminal. The face image data includes face features.

[0096] In this embodiment of the application, when the first terminal fails to recognize the face of the user to be recognized, the first terminal will send the facial feature data of the user to be recognized that it has collected to the scheduling module, such as face images, extracted facial features, and other data.

[0097] In step S130, the facial feature is compared with all registered facial features to obtain N registered facial feature codes that meet preset matching conditions;

[0098] In the embodiments of this application, N is a natural number, such as 1, 10, 50, 100, etc., which can be set according to the actual situation.

[0099] In this embodiment of the application, the registered facial feature code can be used to identify the same user, that is, the facial feature code of the same registered user is the same. Specifically, the registered facial feature code can be encoded by the identity registration platform according to the preset encoding rules after the user registers his or her identity, such as 0001, 0002, etc.

[0100] In this embodiment of the application, comparing the facial feature with all registered facial features to obtain N registered facial feature codes that meet preset matching conditions includes:

[0101] Calculate the comprehensive matching score between the stated facial feature and all registered facial features;

[0102] The comprehensive matching score is compared with a preset score threshold to obtain registered face features whose comprehensive matching score is greater than the preset score threshold.

[0103] Among the registered face features whose matching scores are greater than the preset score threshold, N registered face feature codes are obtained in descending order of the comprehensive matching scores.

[0104] In this embodiment of the application, the facial features may include posture, angle, expression, and different parts, such as eyes and mouth, for comparison, and obtain a matching score for each item. Then, the matching scores of each item are added together to obtain a comprehensive matching score. Furthermore, a weight ratio for each matching item can be set, and the comprehensive matching score is obtained through the matching score and the weight ratio.

[0105] Furthermore, the brightness and darkness of the image can also be matched. Facial features may change under different lighting conditions, so when performing matching, interference from lighting can be eliminated first to improve the accuracy of the matching.

[0106] In this embodiment of the application, after obtaining the comprehensive matching score, it can be compared with a preset score threshold to obtain registered face features with a comprehensive matching score greater than the preset score threshold. Then, among the registered face features with a matching score greater than the preset score threshold, N registered face feature codes are obtained in descending order of the comprehensive matching score as comparison objects.

[0107] In step S140, the second terminal to which the registered face feature code belongs is determined, and the face image data is sent to the second terminal and / or the server for feature comparison.

[0108] In this embodiment, each registered user's facial features have a unique code to identify that user, and each terminal can store different facial features. That is, the codes of the registered facial features stored in each terminal are different. Therefore, after obtaining N registered facial feature codes, they can be allocated to the corresponding second terminals according to the registered facial feature codes. Alternatively, they can be compared with the features sent to the server.

[0109] In step S150, the comparison result sent by the second terminal and / or the server is received and sent to the first terminal so that the first terminal outputs the final face recognition result based on the comparison result.

[0110] In this embodiment of the application, the comparison result may include a successful comparison or a failed comparison. When the comparison is successful, the face feature code of the successful comparison can be sent to the first terminal and the face recognition is successful. When the comparison fails, the result of the failed comparison can be sent to the first terminal and the face recognition fails is output.

[0111] In one embodiment of this application, before receiving the comparison result sent by the second terminal and / or the server and sending it to the first terminal, the process includes:

[0112] After receiving the comparison results sent by the second terminal and the server, it is determined whether the comparison results are consistent;

[0113] If the comparison results are inconsistent, the facial features are compared again with all registered facial features to obtain updated registered facial feature codes.

[0114] Based on the updated registered facial feature code, a third terminal is identified, and the facial feature data and the registered facial feature code are sent to the third terminal.

[0115] In this embodiment, when sending the face image of the user to be identified to the second terminal and the server for feature comparison, the comparison results fed back by the second terminal and the server can be received. The comparison results fed back by the second terminal and the comparison results fed back by the server can be compared to determine whether they are consistent. If they are consistent, they are sent to the first terminal. If they are inconsistent, it indicates that the judgment is wrong. At this time, the scheduling module can re-compare the face feature with all registered face features to obtain the updated registered face feature code. Then, based on the updated registered face feature, it is sent again to the corresponding third terminal for comparison again through the third terminal.

[0116] The third terminal may be one or more, and it stores the updated registered facial features. The third terminal may contain terminal devices different from those in the second terminal; that is, the second terminal devices may include terminal device A and terminal device B, while the third terminal devices may include terminal device H, terminal device G, or either terminal device H or terminal device B.

[0117] In this embodiment of the application, the second terminal includes multiple terminals, and the step of receiving the comparison results sent by the second terminal and / or the server and sending them to the first terminal includes...

[0118] After receiving multiple comparison results sent by the second terminal, determine whether there is a successful comparison result among the comparison results;

[0119] If the judgment result is yes, then send the successful comparison result and the successfully registered face feature code to the first terminal;

[0120] If the judgment result is negative, a comparison failure result is sent to the first terminal.

[0121] In this embodiment, since the scheduling module acquires N registered facial feature codes and each terminal device stores different registered facial features, there are multiple second terminals providing comparison results. Therefore, if there is a successful comparison result among the multiple comparison results, it indicates that the registration information of the user to be identified exists and the first terminal can successfully identify the user. If there are no identification failures among the multiple comparison results, it indicates that the registration information of the user to be identified was not stored before and the first terminal failed to identify the user.

[0122] In this embodiment, due to the limited storage capacity and computing power of the terminal device, only a portion of the registered facial feature codes are stored. When the first terminal fails to recognize the face, the facial feature data of the user to be recognized can be sent to the scheduling module for comparison with all registered facial features. N registered facial feature codes that meet the preset comparison conditions are obtained and sent to the corresponding terminal and / or server for comparison. The comparison result is then fed back to the first terminal as the final facial recognition result. This eliminates the need to store all registered facial features in each terminal device and enables collaborative facial recognition between different devices. It supports large-capacity facial recognition and effectively solves the problems of insufficient computing power and insufficient running memory.

[0123] It should be understood that the sequence number of each step in the above embodiments does not imply the order of execution. The execution order of each process should be determined by its function and internal logic, and should not constitute any limitation on the implementation process of the embodiments of the present invention.

[0124] In one embodiment, an edge-based collaborative face recognition system is provided, which corresponds one-to-one with the edge-based collaborative face recognition method described in the above embodiments. For example... Figure 5 As shown, the edge-based collaborative face recognition system includes a face recognition unit 10, a face image data acquisition unit 20, a registered face feature encoding acquisition unit 30, a feature comparison unit 40, and a face recognition result output unit 50. Detailed descriptions of each functional module are as follows:

[0125] The face recognition unit 10 is used to perform face recognition on the user to be identified through the first terminal device;

[0126] The face image data acquisition unit 20 is used to send the face image data of the user to be identified to the scheduling module through the first terminal device when the face recognition of the first terminal fails. The face image data includes face features.

[0127] The registered face feature encoding acquisition unit 30 is used to compare the face feature with all registered face features to obtain N registered face feature codes that meet preset matching conditions.

[0128] The feature comparison unit 40 is used to determine the second terminal to which the registered face feature code belongs, and to send the face image data to the second terminal and / or the server for feature comparison.

[0129] The face recognition result output unit 50 is used to receive the comparison result sent by the second terminal and / or the server, and send it to the first terminal so that the first terminal outputs the final face recognition result based on the comparison result.

[0130] In one embodiment, the system further includes a registered facial feature distribution unit, used for:

[0131] According to preset rules, different registered facial features are distributed to multiple pre-deployed terminals so that the terminals can perform facial recognition using the registered facial features.

[0132] In one embodiment, the registered face feature encoding acquisition unit 30 is further configured to:

[0133] Calculate the comprehensive matching score between the stated facial feature and all registered facial features;

[0134] The comprehensive matching score is compared with a preset score threshold to obtain registered face features whose comprehensive matching score is greater than the preset score threshold.

[0135] Among the registered face features whose matching scores are greater than the preset score threshold, N registered face feature codes are obtained in descending order of the comprehensive matching scores.

[0136] In one embodiment, the system further includes: an update unit, configured to:

[0137] After receiving the comparison results sent by the second terminal and the server, it is determined whether the comparison results are consistent;

[0138] If the comparison results are inconsistent, the facial features are compared again with all registered facial features to obtain updated registered facial feature codes.

[0139] Based on the updated registered facial feature code, a third terminal is identified, and the facial feature data and the registered facial feature code are sent to the third terminal.

[0140] In one embodiment, the second terminal includes multiple face recognition result output units 50, used for:

[0141] After receiving multiple comparison results sent by the second terminal, determine whether there is a successful comparison result among the comparison results;

[0142] If the judgment result is yes, then send the successful comparison result and the successfully registered face feature code to the first terminal;

[0143] If the judgment result is negative, a comparison failure result is sent to the first terminal.

[0144] In one embodiment, the face recognition unit 10 is further configured to:

[0145] The facial image data of the user to be identified is obtained through the first terminal;

[0146] Feature extraction is performed on the facial image data to obtain the facial features of the user to be identified;

[0147] The facial features are compared with the registered facial features corresponding to the first terminal;

[0148] When the matching degree between the registered facial features and the facial features reaches a preset threshold, the recognition is successful; otherwise, the recognition fails.

[0149] In one embodiment, the face recognition unit 10 is further configured to:

[0150] Perform face detection on the face image data;

[0151] If a face is detected, facial landmark detection is performed on the face image data;

[0152] Based on the results of facial landmark detection, the facial image data is corrected, and the corrected facial image data is adjusted to a facial image of a preset size.

[0153] Image enhancement is performed on the face image of the preset size;

[0154] The enhanced face image is input into the feature model to obtain the face features.

[0155] In this embodiment, due to the limited storage capacity and computing power of the terminal device, only a portion of the registered facial feature codes are stored. When the first terminal fails to recognize the face, the facial feature data of the user to be recognized can be sent to the scheduling module for comparison with all registered facial features. N registered facial feature codes that meet the preset comparison conditions are obtained and sent to the corresponding terminal and / or server for comparison. The comparison result is then fed back to the first terminal as the final facial recognition result. This eliminates the need to store all registered facial features in each terminal device and enables collaborative facial recognition between different devices. It supports large-capacity facial recognition and effectively solves the problems of insufficient computing power and insufficient running memory.

[0156] For specific limitations regarding the edge-based collaborative face recognition system, please refer to the limitations of the edge-based collaborative face recognition method mentioned above, which will not be repeated here. Each module in the aforementioned edge-based collaborative face recognition system can be implemented entirely or partially through software, hardware, or a combination thereof. These modules can be embedded in or independent of the processor in a computer device, or stored in the memory of a computer device as software, so that the processor can call and execute the corresponding operations of each module.

[0157] In one embodiment, a computer device is provided, which may be a terminal device, and its internal structure diagram may be as follows: Figure 6 As shown, the computer device includes a processor, memory, and network interface connected via a system bus. The processor provides computing and control capabilities. The memory includes a readable storage medium storing computer-readable instructions. The network interface communicates with external terminals via a network connection. When executed by the processor, the computer-readable instructions implement an edge-based collaborative face recognition method. The readable storage medium provided in this embodiment includes both non-volatile and volatile readable storage media.

[0158] A computer device includes a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, characterized in that the processor executes the computer-readable instructions to implement the steps of the above-described end-side collaborative face recognition method.

[0159] One or more readable storage media storing computer-readable instructions, characterized in that the computer-readable instructions, when executed by a processor, implement the steps of the above-described end-side collaborative face recognition method.

[0160] Those skilled in the art will understand that all or part of the processes in the methods of the above embodiments can be implemented by instructing related hardware with computer-readable instructions. These computer-readable instructions can be stored in a non-volatile readable storage medium or a volatile readable storage medium. When executed, these computer-readable instructions can include the processes of the embodiments of the above methods. Any references to memory, storage, databases, or other media used in the embodiments provided in this application can include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM), or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in a variety of forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), dual data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link DRAM (SLDRAM), RAMbus direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM).

[0161] Those skilled in the art will clearly understand that, for the sake of convenience and brevity, the above-described division of functional units and modules is used as an example. In practical applications, the above functions can be assigned to different functional units and modules as needed, that is, the internal structure of the system can be divided into different functional units or modules to complete all or part of the functions described above.

[0162] The above-described embodiments are only used to illustrate the technical solutions of the present invention, and are not intended to limit it. Although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features. Such modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention, and should all be included within the protection scope of the present invention.

Claims

1. A face recognition method based on edge-side collaborative technology, characterized in that, The method includes: The first terminal device performs facial recognition on the user to be identified; When face recognition fails, the first terminal sends the face image data of the user to be identified to the scheduling module, and the face image data includes face features. The process involves comparing the facial feature with all registered facial features to obtain N registered facial feature codes that meet preset matching conditions. This includes: calculating the comprehensive matching score between the facial feature and all registered facial features; comparing the comprehensive matching score with a preset score threshold to obtain registered facial features with a comprehensive matching score greater than the preset score threshold; and obtaining N registered facial feature codes from the registered facial features with a matching score greater than the preset score threshold in descending order of the comprehensive matching score. The second terminal to which the registered facial feature code belongs is determined, and the facial image data is sent to the second terminal and / or the server for feature comparison. The system receives the comparison results sent by the second terminal and / or the server, and sends them to the first terminal so that the first terminal outputs the final face recognition result based on the comparison results.

2. The edge-side collaborative face recognition method as described in claim 1, characterized in that, Before performing facial recognition on the user to be identified through the first terminal device, the following steps are included: According to preset rules, different registered facial features are distributed to multiple pre-deployed terminals so that the terminals can perform facial recognition using the registered facial features.

3. The edge-side collaborative face recognition method as described in claim 1 or 2, characterized in that, Before receiving the comparison result sent by the second terminal and / or the server and sending it to the first terminal, the process includes: After receiving the comparison results sent by the second terminal and the server, it is determined whether the comparison results are consistent; If the comparison results are inconsistent, the facial features are compared again with all registered facial features to obtain updated registered facial feature codes. Based on the updated registered facial feature code, a third terminal is identified, and the facial feature data and the registered facial feature code are sent to the third terminal.

4. The edge-side collaborative face recognition method as described in claim 1 or 2, characterized in that, The second terminal includes multiple terminals, and the step of receiving the comparison results sent by the second terminal and / or the server, and sending them to the first terminal includes... After receiving multiple comparison results sent by the second terminal, determine whether there is a successful comparison result among the comparison results; If the judgment result is yes, then send the successful comparison result and the successfully registered face feature code to the first terminal; If the judgment result is negative, a comparison failure result is sent to the first terminal.

5. The edge-side collaborative face recognition method as described in claim 1 or 2, characterized in that, The step of performing facial recognition on the user to be identified via the first terminal device includes: The user's facial image data is obtained through the first terminal; Feature extraction is performed on the facial image data to obtain the facial features of the user to be identified; The facial features are compared with the registered facial features corresponding to the first terminal; When the matching degree between the registered facial features and the facial features reaches a preset threshold, the recognition is successful; otherwise, the recognition fails.

6. The edge-side collaborative face recognition method as described in claim 5, characterized in that, The step of extracting features from the facial image data to obtain the facial features of the user to be identified includes: Perform face detection on the face image data; If a face is detected, facial landmark detection is performed on the face image data; Based on the results of facial landmark detection, the facial image data is corrected, and the corrected facial image data is adjusted to a facial image of a preset size. Image enhancement is performed on the face image of the preset size; The enhanced face image is input into the feature model to obtain the face features.

7. A face recognition system based on edge-side collaboration, characterized in that, The system includes: A face recognition unit is used to perform face recognition on the user to be identified through a first terminal device; A face image data acquisition unit is used to send the face image data of the user to be identified to the scheduling module through the first terminal device when the face recognition of the first terminal fails. The face image data includes face features. The registered face feature encoding acquisition unit is used to compare the face feature with all registered face features to obtain N registered face feature codes that meet preset matching conditions, including: calculating the comprehensive matching score of the face feature and all registered face features respectively; comparing the comprehensive matching score with a preset score threshold to obtain registered face features with a comprehensive matching score greater than the preset score threshold; and obtaining N registered face feature codes in descending order of the comprehensive matching score among the registered face features with a matching score greater than the preset score threshold. The feature comparison unit is used to determine the second terminal to which the registered face feature code belongs, and to send the face image data to the second terminal and / or the server for feature comparison. A face recognition result output unit is used to receive the comparison result sent by the second terminal and / or the server, and send it to the first terminal so that the first terminal outputs the final face recognition result based on the comparison result.

8. A computer device comprising a memory, a processor, and computer-readable instructions stored in the memory and executable on the processor, characterized in that, When the processor executes the computer-readable instructions, it implements the steps of the end-side collaborative face recognition method as described in any one of claims 1 to 6.

9. One or more readable storage media storing computer-readable instructions, characterized in that, When the computer-readable instructions are executed by the processor, they implement the steps of the edge-side collaborative face recognition method as described in any one of claims 1 to 6.

Images