Identity recognition method and apparatus, device, medium and program product

Abstract

Provided in the embodiments of the present application are an identity recognition method and apparatus, a device, a medium and a program product. The method comprises: acquiring an action sequence of biological features; performing feature extraction processing on each biological feature map of a plurality of biological feature maps to obtain a feature vector of each biological feature map; performing fusion processing on the feature vector of each biological feature map to generate a fused feature vector; and, on the basis of the fused feature vector, performing identity recognition on the biological features so as to obtain a recognition result. Using the embodiments of the present application can extract richer feature information of the biological features, thereby improving the recognition accuracy and reliability of a subject to which the biological features belong.

Description

An identity identification method, device, equipment, medium and program product

[0001] This application claims priority to the Chinese patent application filed with the China Patent Office on September 6, 2023, with application number: 202311149538.4, and application name: "A method, device, equipment, medium and program product for identity identification", the entire content of which is incorporated by reference into this application. Technical Field

[0002] The present application relates to the field of computer technology, and in particular to an identity recognition method, an identity recognition device, a computer device, a computer-readable storage medium, and a computer program product. Background Art

[0003] With the advancement of science and technology, biometric recognition is being applied to various business scenarios to improve the convenience of use. For example, in transportation business scenarios, biometric features such as palm prints or faces can be used for identity recognition and authentication, improving the convenience of the transportation process.

[0004] At present, when the feature recognition system collects the user's biometric features, there are often quality problems in the collected biometric images due to various reasons; such as low clarity, low resolution or incomplete feature collection of the biometric images. This makes it difficult to accurately identify the object to which the biometric features belong based on such low-quality images, thereby reducing the reliability of the recognition results.

[0005] Summary of the Invention

[0006] The embodiments of the present application provide an identity recognition method, apparatus, device, medium, and program product that can extract richer feature information of biometrics and improve the accuracy and reliability of identifying objects to which the biometrics belong.

[0007] In one aspect, an embodiment of the present application provides an identity recognition method, which is performed by a computer device and includes:

[0008] An action sequence for acquiring biometric features; the action sequence includes multiple biometric feature graphs continuously collected for the biometric features, each biometric feature graph containing the collected biometric features; and feature changes between the biometric features contained in two adjacent biometric feature graphs in the multiple biometric feature graphs are continuous;

[0009] Performing feature extraction processing on each of the plurality of biometric feature graphs to obtain a feature vector of each biometric feature graph; the feature vector of any biometric feature graph represents sub-feature information of the biometric feature from a semantic dimension of any biometric feature graph;

[0010] The feature vectors of each biometric feature map are fused to generate a fused feature vector; the fused feature vector represents the comprehensive feature information of the biometric feature from the semantic dimension of the action sequence;

[0011] The biometric feature is identified based on the fused feature vector to obtain an identification result, which indicates the object to which the biometric feature belongs.

[0012] On the other hand, an embodiment of the present application provides an identity recognition device, which is mounted on a computer device and includes:

[0013] An acquisition unit configured to acquire an action sequence of a biometric feature; the action sequence includes a plurality of biometric feature graphs continuously acquired for the biometric feature, each biometric feature graph including the acquired biometric feature; and feature changes between two adjacent biometric feature graphs in the plurality of biometric feature graphs are continuous;

[0014] a processing unit configured to perform feature extraction processing on each of the plurality of biometric feature maps to obtain a feature vector for each biometric feature map; the feature vector of any biometric feature map representing sub-feature information of the biometric feature from a semantic dimension of any biometric feature map;

[0015] The processing unit is further configured to fuse the feature vectors of each biometric feature image to generate a fused feature vector; the fused feature vector represents comprehensive feature information of the biometric feature from the semantic dimension of the action sequence;

[0016] The processing unit is further configured to perform identity recognition on the biometric feature based on the fused feature vector to obtain a recognition result, wherein the recognition result indicates the object to which the biometric feature belongs.

[0017] On the other hand, an embodiment of the present application provides a computer device, the computer device comprising:

[0018] a processor for loading and executing computer programs;

[0019] A computer-readable storage medium stores a computer program, which, when executed by a processor, implements the above-mentioned identity recognition method.

[0020] On the other hand, an embodiment of the present application provides a computer-readable storage medium, which stores a computer program, and the computer program is suitable for being loaded by a processor and executing the above-mentioned identity recognition method.

[0021] On the other hand, an embodiment of the present application provides a computer program product, which includes a computer program. When the computer program is executed by a processor, it implements the above-mentioned identity recognition method.

[0022] In an embodiment of the present application, continuous collection of the user's biometric characteristics is supported to obtain an action sequence of the biometric characteristics; wherein, the action sequence includes a plurality of biometric characteristic graphs obtained by continuous collection of the biometric characteristics, and in the plurality of biometric characteristic graphs, according to the order of continuous collection, the feature changes between the biometric characteristics contained in any two adjacent biometric characteristic graphs are continuous. In other words, the embodiment of the present application can ensure that the biometric characteristic sequence composed of the plurality of biometric characteristic graphs included in the action sequence in the order of collection is a continuous action process when the user performs the target action (such as palm swiping action / palm swiping behavior), thereby ensuring the continuity between the biometric characteristics included in the plurality of biometric characteristic graphs in the action sequence. Further, feature extraction processing is performed on the plurality of biometric characteristic graphs in the action sequence to obtain a feature vector of each biometric characteristic graph; the feature vector of any biometric characteristic graph expresses the sub-feature information of the biometric characteristic from the semantic dimension of the any biometric characteristic, such as if the biometric characteristic is a palm, and the palm in the biometric characteristic graph is tilted, then the feature vector in the biometric characteristic graph represents the palm print characteristics when the palm is tilted. Then, the feature vectors of each biometric graph are fused to generate a fused feature vector, which represents the comprehensive feature information of the biometric from the semantic dimension of the action sequence. Based on the fused feature information of the biometric, identity recognition is performed to obtain an identification result indicating the object to which the biometric belongs; since the fused feature information of the biometric is obtained by analyzing the various dimensions of the biometrics included in different biometric graphs, this effectively improves the accuracy of identity recognition by relying on the comprehensiveness of the comprehensive feature information. It can be seen from the above scheme that the embodiment of the present application provides a new biometric collection and recognition scheme, which can continuously collect multiple biometric graphs of the user in the process of performing the target action to form an action sequence, and ensure the continuity of the feature changes between the biometrics in adjacent biometric graphs in the action sequence; in this way, based on a plurality of continuous, comprehensive and rich biometric graphs, the comprehensive feature information of the biometric is obtained, and thus the comprehensive feature information is used for identity recognition. Compared with identity recognition based on the feature vector of a single biometric graph, the accuracy and reliability of the biometrics used for identity recognition are effectively improved, thereby improving the accuracy of identity recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0023] FIG1 is a schematic structural diagram of a feature recognition system provided by an exemplary embodiment of the present application;

[0024] FIG2 is a schematic diagram of the architecture of a business system provided by an exemplary embodiment of the present application;

[0025] FIG3 is a flow chart of an identity recognition method provided by an exemplary embodiment of the present application;

[0026] FIG4 is a schematic diagram of a video stream for collecting biometric features provided by an exemplary embodiment of the present application;

[0027] FIG5 is a schematic diagram of filtering out biometric feature maps constituting an action sequence from a video stream provided by an exemplary embodiment of the present application;

[0028] FIG6 is a schematic diagram of a process of feature extraction and feature fusion provided by an exemplary embodiment of the present application;

[0029] FIG7 is a flow chart of another identity recognition method provided by an exemplary embodiment of the present application;

[0030] FIG8 is a schematic diagram of a process for optimizing image quality for a biometric feature image according to an exemplary embodiment of the present application;

[0031] FIG9 is a schematic structural diagram of an identity recognition device provided by an exemplary embodiment of the present application;

[0032] FIG10 is a schematic structural diagram of a computer device provided by an exemplary embodiment of the present application. DETAILED DESCRIPTION

[0033] The embodiments of the present application relate to biometric recognition technology; biometric recognition technology refers to technology that uses biological characteristics inherent to living beings to perform identity recognition or authentication through computer equipment. Among them, the biological characteristics inherent to a living being can refer to a biological characteristic on the living being's body (such as a body part on the living being's body), and the biological characteristics can include but are not limited to: palm prints, fingerprints, faces, irises, and retinas; each living being's biological characteristics are unique, so the identity of the living being can be recognized based on the biological characteristics. For example, palm print recognition refers to an identification technology that uses the palm print characteristics of the living being's palm (or sole) to realize identity recognition; taking the palm of a living being as an example, the palm print of the living being's palm refers to the palm characteristics from the end of the finger to the wrist. Among them, there are many features in the palm characteristics of a living being that can be used for identity recognition, such as the main line of the palm print, the wrinkles of the palm print, the texture of the palm print, or the bifurcation point of the palm print displayed in the palm to realize palm print recognition. For example, iris recognition refers to an identification technology that uses the iris in the eye (i.e., the flat, circular membrane in the middle layer of the eyeball wall) to identify the individual; this is mainly due to the fact that the iris remains unchanged throughout life, which determines the uniqueness of the iris characteristics.

[0034] It is worth noting that the embodiments of this application do not specify whether the biometric to which the biometric belongs is an animal or a human. In actual applications, the type of organism to which the biometric belongs varies depending on the application field of the biometric recognition technology. For example, in the field of wild animal recognition, the organisms involved in the embodiments of this application are animals, and in this case, the biometrics involved in the embodiments of this application belong to animals. For ease of explanation, the embodiments of this application use the biometric of a palm print and the biometric to which the biometric belongs as a human as an example to introduce the subsequent content, which is specially explained here.

[0035] Furthermore, with the development of biometric recognition technology, non-invasive biometric recognition technologies have also emerged. Non-invasive biometric recognition means that the entire process of collecting and identifying a living being's biometrics is contactless, even imperceptible or silent. Non-contact biometric collection and recognition methods require lower performance from the collection and recognition equipment and do not require direct contact with the living being. They can achieve rapid or multi-person collection and recognition in a short period of time, resulting in high recognition efficiency. However, practical applications have shown that non-contact biometric collection and recognition methods can result in poor quality of collected biometrics for various reasons, significantly reducing the accuracy and reliability of identity recognition. For example, in palmprint recognition, if a user's palm print is collected from a distance, the long distance can result in low resolution and clarity of the captured palmprint image. Therefore, identity recognition based on low-resolution and low-definition palmprint images inevitably results in poor reliability.

[0036] To improve the reliability of biometrics and ensure the accuracy of identity recognition, the present invention provides an identity recognition solution that combines technologies such as continuous video stream acquisition and feature fusion to improve the accuracy and reliability of biometric recognition. Specifically, the general process of this solution may include:

[0037] Acquire a sequence of biometric actions; the action sequence is automatically acquired by the acquisition device when the user performs a target action in front of the acquisition device, such as when the user performs a palm-swiping action in front of the palmprint acquisition device, the palmprint acquisition device automatically acquires the continuous palm-swiping action, and the action sequence includes multiple biometric maps obtained by continuous acquisition of biometrics, and the feature changes between the biometric features contained in any two adjacent biometric maps in the multiple biometric maps collected continuously are continuous, that is, the biometric changes presented by the two adjacent biometric maps are orderly (according to the execution order of the target action). Perform feature extraction processing on each of the multiple biometric maps included in the action sequence to obtain a feature vector for each biometric map; the feature vector of any biometric map represents the sub-feature information of the biometric from the semantic dimension of the biometric map. Fuse the feature vectors of each biometric map in the multiple biometric maps to generate a fused feature vector; compared with the feature vector of a single biometric map, the fused feature vector represents the comprehensive feature information of the biometric from the semantic dimension of the action sequence. Perform biometric identity recognition based on the comprehensive fused feature information to obtain a highly accurate recognition result.

[0038] Thus, the embodiment of the present application can continuously collect the continuous action process of the user's biometrics performing the target action, and obtain multiple continuously collected biometric feature maps, each of which contains biometric features, and the feature changes between the biometric features in adjacent biometric feature maps in the multiple biometric feature maps are continuous; in this way, the feature complementarity (or feature integrity, continuity of feature changes, etc.) between the multiple continuously collected biometric feature maps is utilized to superimpose a richer and more comprehensive fusion feature vector of the biometric features, and the fusion feature vector is used for identity recognition. Compared with identity recognition based on the feature vector of a single biometric feature map, the accuracy and reliability of the biometric features are effectively improved, so that when reliable biometric features are used for identity recognition, the accuracy of identity recognition can be greatly improved. In addition, the recognition scheme provided by the embodiment of the present application is not limited by the distance between the user and the collection device. Even if the user's biometric features are collected from a distance, comprehensive feature information of the biometric features can be obtained by relying on video stream collection, effectively filling the technical gap in the field of long-distance feature collection, providing effective support for the application of biometric recognition technology, and providing reliable support for biometric recognition applications in fields such as security authentication, bringing a higher level of security and convenience to society.

[0039] Based on the above brief introduction to the identity recognition solution provided in the embodiment of this application, the following points need to be explained:

[0040] (1) The identity recognition solution provided in the embodiments of the present application can be deployed in a feature recognition system (or platform) with biometric feature collection and recognition functions.

[0041] Among them, a feature recognition system can refer to an application that supports the collection and recognition of biometric features. An application can be a computer program designed to perform one or more specific tasks. By categorizing applications according to different dimensions (such as the application's operating mode and functionality), the same application can be classified according to different dimensions. For example, based on the application's operating mode, applications may include but are not limited to: clients installed on the terminal, mini-programs (subprograms of the client) that can be used without downloading and installing, and World Wide Web (web) applications opened through a browser. Another example is based on the application's functional type, which may include but is not limited to: instant messaging (IM) applications, content interaction applications, etc. IM applications refer to internet-based applications that enable instant messaging and social interaction. IM applications may include but are not limited to social applications with communication functions, map applications with social interaction functions, and game applications. Content interaction applications refer to applications that enable content interaction, such as online banking, sharing platforms, personal spaces, and news applications.

[0042] Furthermore, the feature recognition system can also be a plug-in (or function) included in the application, which has the function of biometric feature collection and recognition. For example, if the application is an instant messaging application, the feature recognition system can be a biometric feature plug-in included in the instant messaging application. If a user needs to collect and recognize biometric features during a conversation using the instant messaging application, the biometric feature collection and recognition plug-in can be directly called from the instant messaging application to implement the biometric feature collection and recognition function. In this way, users (such as any person using the instant messaging application) can also perform biometric feature collection and recognition functions while using the instant messaging application for social interaction, without having to jump to another application (such as jumping from the instant messaging application to an independent biometric feature collection and recognition application).

[0043] (2) The identification solution provided by the embodiment of the present application can be applied to any business scenario that requires biometric collection and identification. The fields involved in the business scenario may include but are not limited to: security field, transportation field, education field, payment field and personal identity authentication scenario, etc. Taking the biometric image as the palm image of the palm of the object performing the target action, the biometric features contained in the biometric image are palm print features (or simply palm prints), and the business scenario is a long-distance palm swiping scenario as an example, exemplary business scenarios in different fields may include but are not limited to:

[0044] Optionally, business scenarios in the security field may include security access control scenarios. In the security access control scenario, users who are waiting to enter the company can perform palm swiping behavior at a fixed location at a distance or while walking. At this time, the acquisition device (i.e., a computer device deployed with the identity recognition solution (or the feature recognition system deployed by the identity recognition solution) provided by the embodiment of the present application) can continuously capture the palm swiping behavior to obtain a video stream, which includes multiple video frames obtained by the acquisition device continuously capturing the palm swiping behavior, and one video frame corresponds to one biometric feature map. Feature extraction and feature fusion are performed on the multiple continuous biometric feature maps included in the video stream, and based on the fused feature information after feature fusion, the user is allowed to enter the company only when it is determined that the collected palm print is the palm print of an employee within the company.

[0045] Optionally, business scenarios in the transportation sector may include ticket checking scenarios. In ticket checking scenarios, users waiting to enter a station can scan their palms while standing at a fixed location or walking. The user is allowed to enter the station only when the collection device detects that the user whose palm print is collected has purchased a ticket and the ticket has not been used.

[0046] Optionally, business scenarios in the education field may include clocking in. In this scenario, a teacher can scan their palms from a distance before class, either standing still or walking around. The collection device can then associate and record the collected palm prints with the teacher. Specifically, the collected palm prints identify the teacher to whom the palm prints belong, associate the teacher with the palm print activity, and record the teacher's class behavior in a timely manner.

[0047] Optionally, business scenarios in the payment field include offline transfer scenarios. In offline transfer scenarios, consumers can perform palm swiping at a fixed location or while walking. After the collection device collects the consumer's palm print, it can find the consumer's account in the database based on the palm print, deduct the amount of resources corresponding to the goods purchased by the consumer from the resource pool corresponding to the consumer's account, and add the amount of resources corresponding to the goods purchased by the consumer to the resource pool of the store (i.e., the user to be paid) to achieve secure payment.

[0048] It should be understood that, considering that the process and principle of implementing the collection and identification of biometrics in various biometric identification scenarios (or business scenarios) using the identity recognition scheme provided by the embodiment of the present application are similar, it is just that the types of biometrics involved in different identification scenarios may be different. Therefore, the above is only an explanation of the general collection and identification principles in various business scenarios. The relevant descriptions of each business scenario will not limit the embodiments of the present application, and are specifically explained here. For example, business scenarios in the security field can also include detection scenarios; if the video stream collected by the acquisition device in the detection scenario is often low-resolution, if feature recognition is performed based on only a single biometric image in the video stream, the accuracy and reliability of the recognition will be low; therefore, the recognition scheme provided by the embodiment of the present application can be used to fuse the feature vectors of multiple biometric images to obtain a fused feature vector, so that it is easier to identify the user's identity based on a comprehensive and comprehensive fused feature vector.

[0049] It can be seen that compared with the traditional method that requires users to constantly adjust the distance between themselves and the collection device, and adjust the collection angle after the position is fixed so that the collection device can collect a biometric image with better quality; the identity recognition solution provided by the embodiment of the present application does not require the user to remain still during the collection and recognition of biometrics, nor does it limit the distance between the user and the collection device. On the basis of ensuring the accuracy and reliability of biometric recognition, it effectively improves the speed and efficiency of biometric collection and recognition.

[0050] (3) The application structure of the feature recognition system involved in the embodiment of the present application may include but is not limited to four parts. As shown in Figure 1, they are: a biometric feature acquisition subsystem, a feature processing subsystem, a feature extraction and fusion subsystem, and a biometric feature recognition subsystem; wherein:

[0051] ① The biometric acquisition subsystem has a video stream multi-feature map acquisition function. The video stream multi-feature map acquisition function refers to: the technology of continuously acquiring multiple biometric feature maps in the form of video streams; acquiring more palmprint information through continuous acquisition to improve acquisition efficiency and accuracy. The biometric acquisition subsystem includes: a continuous feature map capture module based on video streams and a first image storage module. The continuous feature map capture module is designed to use a special action camera to continuously acquire a video stream of the user's palm swiping behavior, that is, the technology of continuously acquiring multiple palmprint images through video streams, and acquiring multiple biometric feature maps during the palm swiping action through continuous acquisition. The multiple biometric feature maps include rich palmprint information, which improves the efficiency and accuracy of palmprint acquisition. The first image storage module is mainly used to store continuous biometric feature maps in the video stream.

[0052] The feature processing subsystem includes an optimization module and a secondary image storage module. The optimization module optimizes low-resolution images within a series of biometric feature maps, such as using a super-resolution algorithm. The secondary image storage module stores the optimized biometric feature maps. Super-resolution technology improves image resolution through image processing algorithms, reconstructing clearer, more detailed, and high-resolution images from low-resolution images, thereby improving image quality and recognition accuracy.

[0053] ③ The feature extraction and fusion subsystem includes a feature extraction module, a feature storage module, and a feature fusion module. The feature extraction module is used to extract features from each biometric feature map. The feature storage module is used to store the feature vectors of each biometric feature map after feature extraction. The feature fusion module has the function of feature fusion. Feature fusion (or fusion processing) refers to the fusion of feature vectors from multiple biometric features to form a more comprehensive and accurate feature representation; in other words, the feature fusion module is used to fuse the feature vectors of multiple biometric features to obtain more complete and rich comprehensive features. By comprehensively utilizing the features of multiple biometric features, the performance and robustness of the feature recognition system can be improved, and the ability to identify and distinguish individuals can be enhanced.

[0054] ④ The biometric recognition subsystem includes biometric recognition technology; for example, palmprint recognition technology uses features such as palmprint morphology and texture to identify individuals. By analyzing and comparing the features of palmprint images, it confirms and authenticates individual identities, demonstrating uniqueness, stability, and high reliability. The biometric recognition subsystem includes a biometric recognition module, which is primarily used to perform palmprint recognition on the fused feature vector to obtain the recognition result.

[0055] The following is an overview of biometric identification using each module in the feature recognition system's application architecture: When there's a need to collect and identify an object's biometrics, the biometrics of the user (i.e., the object) are continuously collected through the biometric collection subsystem (specifically, the continuous feature graph capture module deployed within the subsystem) to capture a video stream. The biometric collection subsystem (specifically, the first graph storage module deployed within the subsystem) caches multiple continuous biometric graphs in the video stream. The feature processing subsystem (specifically, the optimization module deployed within the subsystem) optimizes the collected biometric graphs, such as using a super-resolution algorithm. The optimized biometric graphs are then stored through the feature processing subsystem (specifically, the second graph storage module deployed within the subsystem). The feature extraction and fusion subsystem (specifically, the feature extraction module deployed within this subsystem) extracts features from each optimized biometric image to obtain a feature vector for each biometric image. Furthermore, the feature extraction and fusion subsystem (specifically, the feature storage module deployed within this subsystem) caches the feature vectors for each biometric image. Furthermore, the feature extraction and fusion subsystem (specifically, the feature fusion module deployed within this subsystem) fuses the feature vectors of each biometric image to generate a fused feature vector. The palmprint recognition subsystem (specifically, the biometric recognition module deployed within this subsystem) recognizes the biometric based on the fused feature vector and obtains a recognition result indicating the object to which the biometric belongs.

[0056] It should be noted that the modules included in the feature recognition system are not limited to the modules given above. For example, a continuity judgment module is also deployed in the feature acquisition subsystem. The continuity judgment module has the function of initializing the video stream. The initialization processing may include but is not limited to: judging the continuity between the biometric features included in the adjacent biometric feature graphs included in the video stream (such as judging whether the current biometric feature graph is the next biometric feature graph of the previous biometric feature graph in the process of the object performing a coherent target action) to ensure the continuity between the biometric features contained in multiple biometric feature graphs, so that the continuity can be used to obtain comprehensive integrated feature information of the biometric features; and archiving the biometric feature graphs after the continuity judgment according to the sequence number of the action sequence. Of course, the function of the continuity judgment module can also be incorporated into the biometric feature acquisition subsystem given above, or incorporated into the feature processing subsystem; the embodiment of the present application does not limit whether the continuity judgment module exists alone or whether the function of the continuity judgment module is incorporated into other modules.

[0057] (4) The identity recognition solution provided in the embodiment of the present application can be executed by a computer device. The embodiment of the present application does not limit the number and type of computer devices. Figure 2 is a schematic diagram of the architecture of a business system provided by an exemplary embodiment of the present application; as shown in Figure 2, the business system includes computer devices 201 and computer devices 202. The embodiment of the present application does not limit the number and naming of computer devices 201 and computer devices 202. Among them:

[0058] ① Computer device 201 may refer to a device with a biometric feature collection function. Specifically, the biometric feature collection subsystem shown in Figure 1 is deployed in the computer device 201. The computer device may be a terminal device that directly interacts with the user; the terminal device may include but is not limited to: smartphones (such as smartphones that deploy the Android system or smartphones that deploy the Internetworking Operating System (IOS)), tablet computers, portable personal computers, mobile Internet devices (MIDs), vehicle-mounted devices, head-mounted devices, and other devices. The embodiments of this application do not limit the type of terminal device, which is explained here.

[0059] ② Computer device 202 is a server corresponding to computer device 201, and is used to interact with computer device 201 to provide computing and application service support for the video stream sent by computer device 201. The server can be an independent physical server, or a server cluster or distributed system composed of multiple physical servers. It can also be a cloud server that provides basic cloud computing services such as cloud services, cloud databases, cloud computing, cloud functions, cloud storage, network services, cloud communications, middleware services, domain name services, security services, content delivery networks (CDNs), and big data and artificial intelligence platforms. Furthermore, a feature processing subsystem and a feature extraction and fusion subsystem are deployed in the server; in this way, after receiving the video stream collected by computer device 201, the server can use the feature processing subsystem to pre-process the video stream, so that the pre-processed video stream is more suitable for subsequent operations, thereby improving the quality of the pre-processed biometric feature map. The feature extraction and fusion subsystem in the server is used to perform feature extraction processing on the preprocessed biometric feature map to obtain the feature vector of the biometric feature map, and to fuse the feature vectors of multiple biometric feature maps to generate a fused feature vector that can comprehensively characterize the characteristics of the biometric features.

[0060] Furthermore, the server also includes a biometric recognition subsystem and a database 203 (or a biometric dataset). The database 203 can be used to store candidate feature information of biometric features of different objects (i.e., users). Thus, after obtaining the fused feature information of the biometric features, the server can use the biometric recognition subsystem to perform recognition processing on the biometric features. Specifically, the server uses the fused feature vector of the biometric features to find candidate feature information from the database 203 that matches (or is identical to) the comprehensive feature information indicated by the fused feature information of the biometric features, thereby treating the object to which the candidate feature information belongs as the object to which the biometric features to be identified belong, thereby achieving the purpose of identifying the object. Furthermore, if the computer device 201 needs to receive feedback information from the computer device 202 before allowing the user to pass (such as in a security access control scenario), then after obtaining the recognition result, the computer device 202 will also return feedback information to the computer device 201 based on the recognition result. The feedback information instructs the computer device 201 to respond to the user's palm-swiping behavior (or other behavior) (such as opening a door or prompting payment completion, etc.).

[0061] It should be understood that the modules in the feature recognition system can be deployed in the computer device 201 or the computer device 202; for example, all modules are deployed in the computer device 201 or the computer device 202, or some modules are deployed in the computer device 201 and the other modules are deployed in the computer device 202. Depending on the deployment location of the modules in the feature recognition system, the execution subject of the identity recognition scheme provided in the embodiment of the present application is different; for example, when all modules are deployed in the computer device 201 (or the computer device 202), the execution subject of the identity recognition scheme is the computer device 201 (or the computer device 202); for another example, when some modules are deployed in the computer device 201 and the other modules are deployed in the computer device 202, the execution subjects of the identity recognition scheme are the computer device 201 and the computer device 202. FIG2 illustrates an example in which the identity recognition scheme is jointly executed by computer device 201 and computer device 202. In actual applications, when computer device 201 has the function of biometric identification, that is, when all modules in the feature recognition system are deployed on computer device 201, the system shown in FIG2 may include only computer device 201. In this case, the identity recognition scheme provided by the embodiment of the present application may be executed only by computer device 201. Similarly, when all modules in the feature recognition system are deployed on computer device 202, the system shown in FIG2 includes computer device 201 and computer device 202. In this case, computer device 201 only performs video stream acquisition and transmits the acquired video stream to computer device 202, and computer device 202 performs identity recognition based on the video stream.

[0062] (5) The collection and processing of relevant data in the embodiments of this application should be strictly in accordance with the requirements of relevant laws and regulations. The acquisition of personal information must be subject to the knowledge or consent of the individual subject (or the legal basis for obtaining the information), and subsequent data use and processing must be carried out within the scope of authorization of laws and regulations and the subject of personal information. For example, when the embodiments of this application are applied to specific products or technologies, such as when collecting the user's biometrics, the user's permission or consent must be obtained, and the collection, use and processing of relevant data (such as the collection and release of the barrage posted by the subject, etc.) must comply with the relevant laws, regulations and standards of the relevant region.

[0063] Based on the identity recognition scheme described above, the embodiment of the present application proposes a more detailed identity recognition method. The identity recognition method proposed in the embodiment of the present application will be described in detail below with reference to the accompanying drawings.

[0064] FIG3 shows a flow chart of an identity recognition method provided by an exemplary embodiment of the present application. The identity recognition method shown in FIG3 can be executed by a computer device, and the identity recognition method may include but is not limited to steps S301-S304:

[0065] S301: Acquire an action sequence of biometric features.

[0066] The action sequence of a biometric feature includes: a plurality of biometric feature graphs obtained by continuously collecting the biometric feature; in other words, the action sequence is a sequence obtained by arranging the plurality of biometric feature graphs in the order in which the biometric feature graphs are collected. The action or style of the biometric feature included in each biometric feature graph is different. For example, if the biometric feature is a palm, then the form of the biometric feature contained in each of the plurality of biometric feature graphs collected continuously in the process of the palm performing the target action is different. Among them, the feature changes between the biometric features contained in any two adjacent biometric feature graphs in the plurality of biometric feature graphs included in the action sequence are continuous; that is, from the perspective of the feature changes between the biometric features contained in the plurality of biometric feature graphs in the action sequence, it can be seen that the changes in the biometric features presented in the entire action sequence are continuous, and this continuity matches the action sequence of the biometric feature performing a coherent target action (such as a palm swiping action). For example, if an action sequence includes an adjacent first biometric feature image and a second biometric feature image, and the first biometric feature image contains a first biometric feature presenting a first action, and the second biometric feature image contains a second biometric feature presenting a second action, then the feature change between the first biometric feature and the second biometric feature (or the change in action from the first action to the second action) is continuous, and this continuity is reflected in the coherence of the action (such as the coherent action process of the palm from clenching a fist to opening it).

[0067] Relying on the continuity (or coherence) of feature changes between the biometric features contained in the multiple biometric feature graphs included in the action sequence, a more comprehensive and reliable biometric fusion feature vector is obtained based on the feature complementarity of the biometric features contained in the multiple consecutive biometric feature graphs. Thus, identity recognition based on this fusion feature vector can improve the accuracy of identity recognition. Taking palm prints (or palm print features) as an example, the continuity of feature changes between the palm prints contained in the multiple biometric feature graphs included in the action sequence can be simply understood as follows: according to the palm print pattern that appears within the capture range of the collection device (i.e., the environmental range that the collection device can capture) when the user performs the palm swiping behavior, the collection device will first capture the fingertips of the palm, then the fingers of the palm, then the partial palm, until the complete palm is captured. Therefore, during the user's palm swiping behavior, the palm print feature changes presented by the multiple consecutively collected biometric feature graphs should be a continuous process from part to whole.

[0068] In actual applications, if there is a need to perform biometric identification on a user in a business scenario, the user's biometric action sequence can be obtained to implement subsequent feature extraction, feature fusion, and feature recognition based on the action sequence to accurately identify the user's identity. The specific implementation process of obtaining the user's biometric action sequence may include but is not limited to:

[0069] (1) When the acquisition device is in the on state, the acquisition device continuously acquires the biometric features of the target action being performed in the business scenario, and obtains a video stream about the biometric features. At this time, the acquired video stream includes multiple initial biometric feature images acquired continuously, and each initial biometric feature image may include all or part of the features of the biometric features (such as a complete palm or a partial palm (such as only the finger part)). In other words, when the acquisition device is in the on state, the process of the acquisition device acquiring the video stream about the biometric features is similar to the process of continuously shooting the biometric features performing the target action. The acquired video stream reflects the continuous process or coherent process of the user's biometric features performing the target action.

[0070] Taking the user's biometric palm print as an example, when the acquisition device is in an on state, the continuous process of the user performing a target action (such as a palm swiping action, or palm swiping behavior) is captured to obtain a video stream of the user's palm print. As shown in Figure 4, when the acquisition device is in an on state, as the user performs the palm swiping behavior, the user's finger first appears within the acquisition range of the acquisition device. The acquisition device then first captures an initial biometric map of the fingertip containing the palm; as the user continues to perform the palm swiping behavior, the acquisition device captures an initial biometric map of the finger containing the palm; as the user continues to perform the palm swiping behavior, the acquisition device captures an initial biometric map containing part of the palm; and so on, until the acquisition device meets the stop acquisition condition, the acquisition device captures N initial biometric maps, where N is an integer greater than 3 (in other cases, N can be an integer greater than 1). Then, it is determined that the video stream captured by the acquisition device for the user's palm print includes N initial biometric maps.

[0071] The embodiments of the present application do not limit the specific content of the conditions for the collection device to stop collection. Optionally, the collection device starts timing from the time the user's biometric characteristics are detected. After continuously collecting the biometric characteristics for a target period of time, the collection device automatically changes from an on state to a off state, thereby stopping collection. For example, if the target period of time is 10 seconds, when the collection device detects that the user's biometric characteristics appear within the collection range, collection begins and continues for 10 seconds. Even if the user's biometric characteristics leave the collection range within the 10 seconds, the collection device will continue to collect. Optionally, when the collection device detects that the user's biometric characteristics fall into the collection range, it begins to collect the biometric characteristics until the biometric characteristics leave the collection range, at which point the collection device stops collecting.

[0072] It can be seen from this that in the video stream obtained by the acquisition device continuously acquiring the biometric features of the target action being performed, the feature changes between the biometric features contained in adjacent different initial biometric maps in multiple initial biometric maps may be relatively large. This will result in the feature changes not being continuous and complementary, and thus when subsequent feature fusion is performed based on the features that are not continuous and complementary, the reliability of the fused feature vector obtained is low. For example, if a user quickly changes from spreading their palms to clenching their palms (such as making a fist), the feature changes between the biometric features contained in the two adjacent initial biometric maps will be too large, resulting in poor subsequent feature fusion effects. For another example, when collecting the user's palm prints, if the user shakes their palms quickly, the feature changes between the biometric features contained in the two adjacent initial biometric maps will also be too large, resulting in poor subsequent feature fusion effects.

[0073] (2) In order to separate multiple biometric feature graphs with strong continuity from the video stream (i.e., multiple initial biometric feature graphs screened from the video stream) to form an action sequence, the continuity and reliability of the feature changes between the biometric features contained in the multiple biometric feature graphs in the action sequence are improved. After collecting the video stream of the biometric features, the embodiment of the present application preprocesses the multiple initial biometric feature graphs included in the video stream to obtain a biometric action sequence; the feature changes between the biometric features contained in the multiple biometric feature graphs included in the action sequence have strong continuity and complementarity.

[0074] The specific implementation logic for preprocessing the video stream to obtain a highly continuous action sequence may include, but is not limited to: in a business scenario, after continuously capturing the target action performed by the user using a dedicated action camera to obtain a video stream, obtaining the feature change requirements of the business scenario. The feature change requirements specify the requirements that must be met for the feature changes between the current initial biometric feature map and the adjacent previous initial biometric feature map in the video stream to be added to the action sequence. According to the order of the initial biometric feature maps included in the video stream, the feature change between the biometric features contained in the two adjacent initial biometric feature maps is sequentially determined to determine whether it meets the feature change requirements of the business scenario. If so, the two adjacent initial biometric feature maps are added to the action sequence; if not, the initial biometric feature map located at the back of the two adjacent initial biometric feature maps, as well as the initial biometric feature map located after the back of the initial biometric feature map in the video stream, are not added to the action sequence.

[0075] For example, taking the case where the video stream includes N initial biometric maps, where N is an integer greater than 1, assume that the feature change requirement of the business scenario can be: the feature change of the biometrics included in two adjacent initial biometric maps in the video stream is continuous. Then, as shown in Figure 5, after continuously collecting the target actions performed by the user using a dedicated action camera to obtain a video stream, it can be determined whether the feature change between the biometrics included in the (i + 1)-th initial biometric map in the video stream and the biometrics included in the i-th initial biometric map in the video stream is continuous; 1 ≤ i < N, and i is an integer. If the feature change between the biometrics included in the (i + 1)-th initial biometric map in the video stream and the biometrics included in the i-th initial biometric map in the video stream is continuous, it indicates that there is action coherence between the biometrics included in the i-th biometric map and the biometrics included in the (i + 1)-th biometric map, that is, the i-th biometric map and the (i + 1)-th biometric map can be used to achieve feature fusion and recognition, then add the (i + 1)-th initial biometric map to the action sequence of the biometrics. And, continue to determine whether there is an (i + 2)-th initial biometric map in the video stream (that is, determine whether the (i + 1)-th initial biometric map is the last one of the N initial biometric maps included in the video stream), if it exists (that is, the (i + 1)-th initial biometric map is not the last one of the N initial biometric maps included in the video stream), then continue to determine whether the biometric change between the biometrics included in the (i + 2)-th initial biometric map and the biometrics included in the (i + 1)-th initial biometric map is continuous, until the candidate biometric map in the video stream is detected; the candidate biometric map includes: the last initial biometric map among the N initial biometric maps included in the video stream, or, among the N initial biometric maps, the initial biometric map whose feature change between the biometrics included in the biometric map and the biometrics included in the previous adjacent initial biometric map is not continuous.

[0076] Conversely, if the feature change between the biometrics included in the (i + 1)-th initial biometric map in the video stream and the biometrics included in the i-th initial biometric map in the video stream is not continuous (as shown in Figure 5, the palm included in the (i + 1)-th initial biometric map is in a clenched state, while the palm included in the i-th initial biometric map is in an open state, then the two are not continuous), it indicates that the biometrics included in the (i + 1)-th biometric map have a large action change relative to the biometrics included in the i-th biometric map, without coherence, and cannot be used to achieve feature fusion and recognition; then determine the action sequence of the biometrics. At this time, the action sequence includes i initial biometric maps in the video stream, and the i initial biometric maps are: the initial biometric maps in the video stream whose acquisition order is before the (i + 1)-th initial biometric map).

[0077] It should be understood that the purpose of preprocessing the video stream is to screen out multiple biometric graphs with strong action continuity between biometric features from the video stream to form an action sequence, thereby ensuring that the feature changes between the biometric features contained in the multiple biometric graphs contained in the action sequence are continuous. Therefore, the embodiments of the present application do not limit the specific algorithm used for preprocessing the video stream. For example, the embodiments of the present application support the use of a deep learning model to determine whether the feature changes between the biometric features contained in two adjacent initial biometric graphs in the video stream are continuous. In other words, the trained deep learning model can learn and understand the action relationship between the biometric features contained in different initial biometric graphs, that is, to determine whether the action presented by the biometric features contained in an initial biometric graph is the next action of the action presented by the biometric features contained in the adjacent previous initial biometric graph. The embodiments of the present application do not limit the model structure and type of the deep learning model. For example, the deep learning model may include but is not limited to: one or more of: Long Short-Term Memory Network (LSTM), Artificial Neural Network (ANN) and Multilayer Perceptron (MLP).

[0078] Furthermore, the computer device 202 in the system shown in FIG2 can deploy the deep learning model; in this implementation, after the computer device 201 captures the video stream through a dedicated action camera, the continuously captured video stream can be transmitted to the computer device 202, and the computer device 202 calls the trained deep learning model to perform continuity judgment on the adjacent initial biometric feature maps included in the video stream. Among them, if the feature changes between the biometric features contained in two adjacent initial biometric feature maps are continuous, then the two adjacent initial biometric feature maps can be archived so as to be added to the action sequence of the biometric feature, ensuring the continuity of the feature changes between the biometric features contained in the multiple biometric feature maps included in the action sequence, as well as the integrity of the biometric features contained in the multiple biometric feature maps (such as the integrity is reflected in the feature changes of the biometric features reflected by the multiple biometric feature maps being the feature changes corresponding to the biometric feature executing the target action). On the other hand, if the feature changes between the biometric features contained in two adjacent initial biometric feature maps are not continuous (e.g., the feature changes between the biometric features contained in the i-th initial biometric feature map and the biometric features contained in the i+1-th initial biometric feature map are not continuous), that is, an action interruption occurs, then the i+1-th initial biometric feature map can be appropriately processed to add the processed i+1-th initial biometric feature map to the interference dataset. The interference dataset includes: an initial biometric feature map in which the biometric feature changes between the biometric features in the video stream and the biometric features contained in the previous adjacent initial biometric feature map are not continuous; this interference dataset can be used as interference data for the deep learning model during the model training process. Thus, on the one hand, the initial biometric feature map corresponding to the interrupted action is appropriately processed (such as labeling, regularization, and dimensionality reduction, etc.) and then cached in the interference dataset, and the interference dataset is used as data for model training. This method can effectively utilize the collected initial biometric feature maps, which is not only beneficial to the interference training of the deep learning model, but also avoids the waste of the collected initial biometric feature maps. On the other hand, the embodiment of the present application uses a deep learning model to preprocess the video stream. With the help of the powerful learning ability of the deep learning model, it can automatically preprocess the video stream, get rid of manual preprocessing and improve the quality and efficiency of preprocessing.

[0079] In summary, the video stream multi-feature map acquisition technology enables real-time and continuous capture of the user's target actions, ensuring the coherence between the initial biometric maps contained in the captured video stream. Furthermore, it supports continuity judgment of feature changes between biometric features contained in adjacent initial biometric maps in the video stream to construct biometric action sequences. This ensures the continuity, complementarity, and integrity of the biometric features contained in the multiple biometric maps within the action sequence, thereby obtaining richer feature information about the user and improving the reliability and authenticity of the biometric features.

[0080] S302: Perform feature extraction processing on each of the multiple biometric feature maps to obtain a feature vector of each biometric feature map.

[0081] S303: Fusing the feature vectors of each of the multiple biometric feature maps to generate a fused feature vector.

[0082] In steps S302-S303, after obtaining the biometric action sequence based on the aforementioned step S301, a continuous feature graph feature extraction and fusion module can be used to extract and fuse features for each biometric graph in the action sequence. This is intended to utilize multiple continuous biometric graphs to extract and fuse biometric features, thereby further improving the accuracy and reliability of biometric recognition. The feature extraction and fusion of the biometric graphs in the action sequence can include two parts: feature extraction for each biometric graph in the action sequence, and feature fusion of feature vectors for multiple biometric graphs. Specifically:

[0083] ① The feature extraction processing for each biometric graph in the action sequence is intended to remove redundant information and noise in the original data in the biometric graph, thereby reducing the dimension of the data, which is conducive to analyzing the key features of the biometrics in the biometric graph (such as color, lines and patterns, etc.) and improving the data processing efficiency. Furthermore, feature extraction can also map the biometric graph into a feature space and express the semantics of the biometric graph in vector form to extract a feature vector representing the semantic features of the biometric graph. The embodiment of the present application does not limit the type of algorithm for feature extraction of biometric graphs. For example, the embodiment of the present application supports training a deep learning model with a feature extraction function, which is capable of extracting features from continuous biometric graphs after training; wherein, the deep learning model with a feature extraction function may include but is not limited to: Recurrent Neural Network (RNN), Convolutional Neural Networks (CNN) and Transformer network, etc.

[0084] Recurrent neural networks are a type of neural network with short-term memory. Neurons in recurrent neural networks can receive information not only from other neurons but also from their own neurons, forming a network structure with loops. Convolutional neural networks are a type of feedforward neural network that incorporates convolutional computations. They are designed based on the translation invariance of image tasks (objects in image recognition have the same meaning in different locations). The Transformer network is a neural network model based on the attention mechanism. The Transformer network abandons the convolutional model structure and achieves good performance solely through the attention mechanism and a feedforward neural network, eliminating the need for a sequence-aligned recurrent architecture. The attention mechanism is a solution proposed to mimic human attention. Simply put, it mimics human attention in quickly filtering desired information from a large amount of information. It is primarily used to address the difficulty of obtaining a reasonable vector representation for long input sequences in time series models. This approach involves retaining the intermediate results of the time series model, learning them with a new model, and associating them with the output to achieve the purpose of information filtering.

[0085] ② Feature fusion (or fusion processing) of feature vectors from multiple biometric maps refers to the process of fusing the feature vectors of multiple biometric maps to form a more comprehensive and accurate fused feature vector. Feature fusion aims to integrate the complete and rich feature information extracted from each biometric map included in the action sequence to obtain a fused feature vector that can represent the important information of all biometric maps. The fused fused feature vector can contain more comprehensive and rich biometric information, improve the reliability and accuracy of the biometrics, and thus enhance the performance of biometric recognition. Methods for fusing feature vectors from multiple biometric maps in an action sequence may include, but are not limited to, feature concatenation or feature weighting. Feature concatenation refers to the direct concatenation of feature vectors from multiple biometric maps. For example, if the dimensions of the feature vectors of two biometric maps are p and q, respectively, the dimension of the concatenated feature vector is p+q. Feature weighting refers to adding weights to the feature vectors of multiple biometric maps to achieve weighted processing of the feature vectors of multiple biometric maps. For example, if the feature vectors of two biometric maps are p and q respectively, then the feature vector after feature weighting can be expressed as z = xp + yq, where z is the fused feature vector, x is the weight value of the feature vector p of one biometric map, and y is the weight value of the feature vector q of the other biometric map.

[0086] In summary, the overall process for extracting and fusing features from multiple biometric graphs in an action sequence by the continuous feature graph feature extraction and fusion module can be roughly illustrated in Figure 6. This process may include: performing feature extraction processing on each of the multiple biometric graphs in the action sequence in a serial or parallel manner (e.g., using a deep learning model with feature extraction capabilities to extract features from each biometric graph), extracting a feature vector for each biometric graph; the feature vector of each biometric graph is used to represent partial feature information of the biometric features contained in the biometric graph based on the semantic dimension of the individual biometric graph. Feature fusion is then performed on the feature vectors of each of the multiple biometric graphs in the action sequence (e.g., using the aforementioned fusion methods such as feature concatenation or feature weighting) to generate a fused feature vector; this fused feature vector utilizes the complementarity (or authenticity, or superposition) between the biometric features contained in the multiple consecutive biometric graphs in the action sequence to represent the comprehensive feature information of the biometric features. Thus, by fusing the feature information of all the biometric graphs in the action sequence represented by the feature vector, the performance and robustness of the feature recognition module are improved, and the ability to identify and authenticate individual identities is enhanced.

[0087] S304: Perform identity recognition on the biometric features based on the fused feature vector to obtain a recognition result.

[0088] After obtaining the fused feature vector of the biometric features based on the aforementioned steps, the identity of the object to which the biometric features belong can be identified based on the fused feature vector to identify the object to which the biometric features belong, thereby realizing the recognition of the individual's identity. The feature recognition module provided in the embodiment of the present application relies on the comparison and matching technology of biometric features, and uses a feature comparison algorithm to process and analyze the input fused feature vector to determine its similarity with the candidate biometric features in the biometric feature dataset (such as the database mentioned above), thereby realizing identity recognition and matching based on biometric features.

[0089] In summary, after the embodiment of the present application obtains a video stream by real-time and continuous acquisition of the target action performed by the user, the video stream can be pre-processed to obtain an action sequence with strong continuity. In this way, by extracting and fusing features of multiple biometric maps in the action sequence, more comprehensive and rich feature information of the biometrics can be obtained, and the perception of individual differences of the subsequent feature recognition model can be enhanced. In detail, ① by fusing the features of multiple consecutive biometric maps, the impact of noise and changes brought by a single biometric map can be reduced, and the robustness of the system can be improved (that is, when some biometric maps are interfered with or of poor quality, the features of other biometric maps can still make up for it and provide reliable identification information). ② The feature fusion of multiple consecutive biometric maps can improve the one-to-one matching performance; specifically, the feature fusion of multiple biometric maps can provide more comprehensive and accurate feature information, better capture the details and uniqueness of the biometrics, and improve the accuracy of subsequent biometric matching. ③ The feature fusion of multiple consecutive biometric feature maps can also improve the one-to-many retrieval effect; specifically, in the one-to-many identity recognition process, the feature fusion of multiple consecutive biometric feature maps can provide richer feature information, enhance the feature recognition module's ability to distinguish biometric features, improve the feature recognition module's retrieval effect in large-scale databases, and reduce the false recognition rate. ④ The continuous feature map acquisition technology based on video streams enables the embodiment of the present application to support dynamic biometric feature recognition; specifically, the feature fusion of multiple consecutive biometric feature maps enables the feature recognition module to recognize dynamic biometric features, that is, when the biometric feature is performing continuous actions, the feature recognition model can capture the temporal change pattern of the biometric feature, thereby achieving accurate recognition of dynamic biometric features.

[0090] FIG7 shows a flow chart of another identity recognition method provided by an exemplary embodiment of the present application. The identity recognition method shown in FIG7 can be executed by a computer device, and the identity recognition method may include but is not limited to steps S701-S705:

[0091] S701: Acquire an action sequence of biometric features.

[0092] It should be noted that the specific implementation process shown in step S701 can refer to the relevant description of the specific implementation process shown in step S301 in the embodiment shown in Figure 3 above, and will not be repeated here.

[0093] As can be seen from the foregoing description, the multiple consecutive biometric graphs included in the biometric action sequence are part of the initial biometric graphs filtered from the video stream. To improve the image quality of the biometric graphs included in the action sequence and increase the accuracy of identity recognition, the embodiments of the present application also support image enhancement processing of the biometric graphs in the action sequence to improve image quality. This is conducive to extracting more realistic and accurate feature information of the biometric based on the high-quality biometric graphs, thereby improving the perception of personal identity.

[0094] The following describes the process of optimizing the quality of biometric images in an action sequence, using Figure 8. As shown in Figure 8, after acquiring a biometric action sequence, image quality testing is performed on each biometric image in the action sequence to obtain a quality test result for each biometric image. For example, resolution testing is performed on the biometric image. The lower the resolution of the biometric image, the lower the clarity of the biometric image; the higher the resolution of the biometric image, the higher the clarity of the biometric image. This embodiment of the application does not limit the specific algorithm and process for image quality testing. Based on the quality test results of the biometric image, it is determined whether the biometric image requires image enhancement processing. Specifically, biometric images from the action sequence whose quality test results do not meet the quality requirements are screened out. For example, biometric images from the action sequence whose resolution falls below a resolution threshold are screened out as biometric images whose quality test results do not meet the quality requirements. Image enhancement processing is performed on the biometric images whose quality test results do not meet the quality requirements to obtain enhanced biometric images. Biometric images whose quality test results meet the quality requirements are directly used for subsequent feature extraction. Image enhancement can be simply understood as the process of making an originally unclear image clearer or emphasizing interesting features in the image, magnifying the differences between features of different objects in the image, improving image quality and enriching the amount of information to enhance image interpretation and recognition. The embodiments of this application do not limit the image enhancement algorithm used for image enhancement processing.

[0095] Exemplarily, the embodiments of the present application support personalized selection of appropriate image enhancement algorithms for different biometric images based on the image processing requirements of each biometric image that requires image enhancement; by matching the image processing requirements of such biometric images and the image enhancement algorithm, the image quality of a single biometric image after image enhancement processing is improved. In a specific implementation, taking any biometric image in an action sequence whose quality detection result does not meet the quality requirements as an example, the process of performing image enhancement processing on the biometric image whose quality detection result does not meet the quality requirements may include, but is not limited to: obtaining the image processing requirements of the target biometric image and selecting an image enhancement algorithm that matches the image processing requirements; performing image enhancement processing on the target biometric image using the image processing algorithm that matches the image processing requirements to obtain an enhanced biometric image after image enhancement processing. Among them, matching the image processing requirements and the image enhancement algorithm may mean that the result obtained by performing image enhancement processing on the biometric image using the image enhancement algorithm can meet the image processing requirements. For example, if the image processing requirement for a target biometric image is to enhance high-frequency information within the target biometric image, image enhancement algorithms that match this requirement may include, but are not limited to, interpolation and learning algorithms within super-resolution algorithms. Super-resolution technology uses image processing algorithms to increase image resolution; it can recover a clearer, more detailed, high-resolution image from a low-resolution image, thereby improving image quality and recognition accuracy. Specifically, super-resolution technology typically involves two key principles: interpolation and learning. Interpolation is a fundamental principle of super-resolution technology; it interpolates pixels in a low-resolution image, inferring missing high-frequency details based on information from neighboring pixels. Common interpolation methods include bilinear interpolation and bicubic interpolation. Learning is a more advanced principle within super-resolution technology; it uses deep learning models to learn the mapping relationship between low-resolution and high-resolution images. Through model training, it learns a more accurate method for recovering details, thereby generating high-quality high-resolution images. Of course, the image enhancement algorithm used to perform image enhancement on the biometric feature image may include not only the super-resolution algorithm but also other algorithms (such as grayscale change algorithm or histogram, etc.), which is not limited to this.

[0096] Furthermore, the enhanced biometric image after image enhancement is used to update the biometric action sequence to obtain an updated target action sequence; the updated target action sequence includes the enhanced biometric image after image enhancement and the biometric image whose quality inspection results meet the quality requirements. Specifically, the biometric image whose quality inspection results do not meet the quality requirements is deleted from the biometric action sequence, and the enhanced biometric image after image enhancement corresponding to the biometric image whose quality inspection results do not meet the quality requirements is inserted into the action sequence according to the arrangement position of the biometric image in the action sequence to obtain the updated target action sequence.

[0097] It can be seen that the embodiment of the present application uses an image enhancement algorithm (such as super-resolution technology) to process the collected biometric image to improve the details and clarity of the image, so that the enhanced biometric image after enhancement processing is clearer and the lost details in the biometric image are restored, which is conducive to the subsequent reference to more accurate and detailed feature information about the biometrics from the high-quality biometric image, thereby improving the recognition performance of the biometrics.

[0098] S702: Perform feature extraction processing on each of the multiple biometric feature maps to obtain a feature vector of each biometric feature map.

[0099] S703: Fusing the feature vectors of each of the multiple biometric feature maps to generate a fused feature vector.

[0100] It should be noted that, for the specific implementation process shown in steps S702-S703, reference can be made to the relevant description of the specific implementation process shown in steps S302-S303 in the embodiment shown in FIG3 , and no further details are given here.

[0101] S704: Obtain a biometric feature dataset, and use a feature comparison algorithm to compare the comprehensive feature information indicated by the fused feature vector with each candidate feature information in the biometric feature dataset to obtain a feature matching result corresponding to each candidate feature information.

[0102] S705: Generate a recognition result based on the feature matching result corresponding to each candidate feature information.

[0103] In steps S704-S705, after obtaining the fused feature vector of multiple biometric feature images in the action sequence based on the aforementioned steps, identity recognition is performed on the biometric feature from feature dimensions such as the morphology and texture of the biometric feature based on the fused feature vector to achieve confirmation and authentication of the individual's identity. Identity recognition of the biometric feature is performed based on the comprehensive feature information of the biometric feature indicated by the fused feature vector, which not only improves the accuracy of the recognition result, but also makes the recognition of the individual's identity unique, stable and highly reliable.

[0104] In the specific implementation: (1) a biometric dataset is first obtained, which includes candidate feature information of biometric features of different objects. In other words, the biometric dataset is a database that pre-stores the real biometric features of a large number of users; thus, when user identification is required, identity recognition and matching can be performed based on the real database.

[0105] (2) Using a feature comparison algorithm, the comprehensive feature information indicated by the fused feature vector is compared with each candidate feature information in the biometric dataset to obtain a feature matching result corresponding to each candidate feature information. The feature comparison algorithm may include, but is not limited to, a trained deep learning model or other pattern recognition algorithm. The following uses a feature comparison algorithm including a similarity algorithm and a distance comparison algorithm as an example to exemplify the process of using the feature comparison algorithm to compare the comprehensive feature information indicated by the fused feature vector with each candidate feature information in the biometric dataset to obtain a feature matching result corresponding to each candidate feature information.

[0106] In a specific implementation, on the one hand, a similarity algorithm can be used to calculate the similarity of the comprehensive feature information indicated by the fused feature vector and each candidate feature information in the biometric dataset based on the similarity dimension of the image content, thereby obtaining a similarity comparison result corresponding to each candidate feature information. On the other hand, a distance comparison algorithm can be used to calculate the distance index of the first distance between the biometric feature corresponding to the fused feature vector and the acquisition device at the time of acquisition, and the second distance corresponding to each candidate feature information in the biometric dataset based on the shooting distance dimension during image acquisition, thereby obtaining a distance index result corresponding to each candidate feature information; the distance index result is used to indicate the distance difference between the first distance and the second distance; the second distance corresponding to the candidate feature information refers to the distance between the biometric feature corresponding to the candidate feature information and the acquisition device at the time of acquisition.

[0107] Based on the similarity comparison result and distance index result corresponding to each candidate feature information, a feature matching result corresponding to each candidate feature information is generated. Specifically, ① when the feature matching result is the weighted value of the similarity comparison result and the distance index result, the feature matching result satisfies the matching conditions including: the weighted value is greater than the weighted threshold. For example, the feature matching result corresponding to candidate feature information 1 = the similarity comparison result corresponding to candidate feature information 1 * 20% + the distance index result corresponding to candidate feature information 1 * 80%. If the feature matching result is greater than the weighted threshold, it is determined that the feature matching result corresponding to candidate feature information 1 meets the matching conditions, that is, there is candidate feature information in the biometric dataset whose feature matching result meets the matching conditions. ② When the feature matching result includes the similarity comparison result and the distance index result, the feature matching result satisfies the matching conditions including: the similarity comparison result is greater than the similarity threshold, and the distance index result is greater than the distance difference threshold. For example, the feature matching results corresponding to candidate feature information 1 include: the similarity comparison result corresponding to candidate feature information 1 and the distance index result corresponding to candidate feature information 1; if the similarity comparison result corresponding to candidate feature information 1 is greater than the similarity threshold, and the distance index result corresponding to candidate feature information 1 is greater than the distance difference threshold, then it is determined that the feature matching result corresponding to candidate feature information 1 meets the matching condition. It can be seen that the embodiment of the present application can combine the two dimensions of similarity comparison result and distance index result to jointly generate the feature matching result corresponding to each candidate feature information. Compared with predicting the feature matching result based on a single-dimensional index result, the accuracy of the feature matching result is improved to a certain extent, thereby improving the accuracy of subsequent identity recognition.

[0108] (3) Generate an identification result based on the feature matching result corresponding to each candidate feature information. According to the feature matching result corresponding to each candidate feature information, it can be determined that the identification result indicates that there is candidate feature information in the biometric data set whose feature matching result satisfies the matching condition, or that the identification result indicates that there is no candidate feature information in the biometric data set whose feature matching result satisfies the matching condition. For example, assuming that the candidate feature information whose feature matching result satisfies the matching condition includes: the value of the feature matching result is greater than the matching threshold 95%, then if the biometric data set includes candidate feature information 1, candidate feature information 2 and candidate feature information 3, and the feature matching result between candidate feature information 1 and the comprehensive feature information indicated by the fused feature vector is 10%, the feature matching result between candidate feature information 2 and the comprehensive feature information indicated by the fused feature vector is 96%, and the feature matching result between candidate feature information 3 and the comprehensive feature information indicated by the fused feature vector is 0.1%, then it is determined that the object to which candidate feature information 2, whose feature matching result is greater than 95%, belongs is the object for identity recognition of the target action being performed this time, and at this time, it is determined that there is candidate feature information in the biometric data set whose feature matching result satisfies the matching condition.

[0109] It is worth noting that when the recognition result indicates that there is candidate feature information in the biometric dataset whose feature matching result meets the matching conditions, the recognition result may also include the object information of the object to which the biometric belongs. The object information included in the recognition result varies depending on the business scenario's requirements for the user's object information. For example: in a transportation ticket purchase scenario, the recognition result needs to include the account information, train information, and transaction information of the object to which the biometric belongs; for another example: in a security access control scenario, the recognition result needs to include the object name of the object to which the biometric belongs. The embodiment of the present application does not limit the type and quantity of object information specifically included in the recognition algorithm.

[0110] In the embodiment of the present application, multiple biometric feature maps of the user during the execution of the target action (such as the palm swiping action / swiping behavior) can be continuously collected, and the comprehensive and rich vector representation of the biometric features based on the analysis of the multiple biometric feature maps can be used for identity recognition. Compared with identity recognition based on the feature vector of a single biometric feature map, the accuracy and reliability of the biometric features used for identity recognition are effectively improved, thereby improving the accuracy of identity recognition. In addition, image enhancement processing is supported for some biometric feature maps in the action sequence whose quality inspection results do not meet the quality requirements; this selective image enhancement processing method for some biometric feature maps can not only save image processing resources, but also ensure high-resolution images of the biometric feature maps in the action sequence, thereby ensuring the reliability of subsequent feature extraction, feature fusion and recognition, thereby improving the recognition performance of personal identity.

[0111] The above describes in detail the method of the embodiment of the present application. In order to facilitate better implementation of the above scheme of the embodiment of the present application, the device of the embodiment of the present application is provided below accordingly.

[0112] FIG9 shows a schematic diagram of the structure of an identity recognition device provided by an exemplary embodiment of the present application; the identity recognition device can be used to perform some or all of the steps in the method embodiments shown in FIG3 or FIG7. Referring to FIG9, the identity recognition device includes the following units:

[0113] An acquisition unit 901 is configured to acquire an action sequence of a biometric feature; the action sequence includes a plurality of biometric feature graphs continuously collected for the biometric feature; and the feature changes between two adjacent biometric feature graphs in the plurality of biometric feature graphs are continuous;

[0114] A processing unit 902 is configured to perform feature extraction processing on each of the plurality of biometric feature maps to obtain a feature vector for each biometric feature map; the feature vector of any biometric feature map represents sub-feature information of the biometric feature from a semantic dimension of any biometric feature map;

[0115] The processing unit 902 is further configured to fuse the feature vectors of each biometric feature image to generate a fused feature vector; the fused feature vector represents comprehensive feature information of the biometric feature from the semantic dimension of the action sequence;

[0116] The processing unit 902 is further configured to perform identity recognition on the biometric feature based on the fused feature vector to obtain a recognition result, where the recognition result indicates the object to which the biometric feature belongs.

[0117] In one implementation, the acquiring unit 901 is configured to acquire a biometric action sequence, specifically to:

[0118] In a business scenario, biometric features of a target action are continuously collected to obtain a video stream of the biometric features; the video stream includes a plurality of continuously collected initial biometric feature maps, each of which includes all or part of the biometric features;

[0119] Preprocessing a plurality of initial biometric feature images included in the video stream to obtain a biometric action sequence;

[0120] The biometric image is a palm image of the object performing the target action, the biometric features included in the biometric image are palm print features, and the business scenario is a long-distance palm scanning scenario.

[0121] In one implementation, the video stream includes N initial biometric maps, where N is an integer greater than 1; the processing unit 902 is specifically configured to preprocess the multiple initial biometric maps included in the video stream to obtain the action sequence of the biometric features, and specifically includes:

[0122] Determine whether the feature change between the biometric features included in the (i + 1)-th initial biometric map in the video stream and the biometric features included in the i-th initial biometric map in the video stream is continuous; 1 ≤ i < N, and i is an integer;

[0123] If the feature change between the biometric features included in the (i + 1)-th initial biometric map and the biometric features included in the i-th initial biometric map in the video stream is continuous, add the (i + 1)-th initial biometric map to the action sequence of the biometric features, and determine whether the (i + 1)-th initial biometric map is the last one among the N initial biometric maps included in the video stream;

[0124] If the (i + 1)-th initial biometric map is not the last one among the N initial biometric maps included in the video stream, continue to determine whether the feature change between the biometric features included in the (i + 2)-th initial biometric map and the biometric features included in the (i + 1)-th initial biometric map is continuous until a candidate biometric map in the video stream is detected;

[0125] Among them, the candidate biometric map includes: the last initial biometric map among the N initial biometric maps; or, among the N initial biometric maps, the initial biometric map whose feature change between the biometric features and the biometric features included in the adjacent previous initial biometric map is not continuous.

[0126] In one implementation, the processing unit 902 is further configured to:

[0127] If the feature change between the biometric features included in the (i + 1)-th initial biometric map and the biometric features included in the i-th initial biometric map in the video stream is not continuous, determine the action sequence of the biometric features; the action sequence includes i initial biometric maps in the video stream, and the i initial biometric maps are: the initial biometric maps whose acquisition order in the video stream is before the (i + 1)-th initial biometric map.

[0128] In one implementation, if the determination of whether the feature change between the biometric features included in the (i + 1)-th initial biometric map and the biometric features included in the i-th initial biometric map is continuous is implemented by using a deep learning model, the processing unit 902 is further configured to:

[0129] Add the (i + 1)-th initial biometric map to the interference data set;

[0130] Among them, the interference data set includes: an initial biometric feature map in which the feature changes between the biometric features in the video stream and the biometric features contained in the previous adjacent initial biometric feature map are not continuous; the interference data set serves as interference data for the deep learning model during the model training process.

[0131] In one implementation, the processing unit 902 is further configured to:

[0132] Performing image quality detection on each biometric feature image in the action sequence to obtain a quality detection result of each biometric feature image;

[0133] Filter out biometric features whose quality inspection results do not meet quality requirements from the action sequence;

[0134] Performing image enhancement processing on the biometric image whose quality inspection result does not meet the quality requirements to obtain an image-enhanced biometric image;

[0135] The enhanced biometric image is used to update the action sequence of the biometrics to obtain an updated target action sequence; the target action sequence includes: the enhanced biometric image after image enhancement and the biometric image whose quality detection result meets the quality requirements.

[0136] In one implementation, a biometric image whose quality inspection result does not meet the quality requirements is represented as a target biometric image; the processing unit 902 is configured to perform image enhancement processing on the biometric image whose quality inspection result does not meet the quality requirements, and to obtain an enhanced biometric image after image enhancement, specifically to:

[0137] Obtain the image processing requirements of the target biometric image and select an image enhancement algorithm that matches the image processing requirements;

[0138] Using an image processing algorithm that matches the image processing requirements, the target biometric image is enhanced to obtain an enhanced biometric image after image enhancement processing;

[0139] Among them, if the image processing requirement is to enhance the high-frequency information in the enhanced biometric feature image, the image enhancement algorithms that match the image processing requirements include: interpolation algorithms and learning algorithms under super-resolution algorithms.

[0140] In one implementation, the processing unit 902 is configured to perform identity recognition on the biometric features based on the fused feature vector, and when obtaining the recognition result, specifically to:

[0141] Acquire a biometric feature dataset, where the biometric feature dataset includes candidate feature information of biometric features of different objects;

[0142] Using the feature comparison algorithm, the comprehensive feature information indicated by the fused feature vector is compared with each candidate feature information in the biometric dataset to obtain the feature matching result corresponding to each candidate feature information;

[0143] Generate recognition results based on the feature matching results corresponding to each candidate feature information;

[0144] The recognition result indicates that candidate feature information exists in the biometric feature data set whose feature matching result satisfies the matching condition, or the recognition result indicates that candidate feature information does not exist in the biometric feature data set whose feature matching result satisfies the matching condition.

[0145] In one implementation, the feature comparison algorithm includes a similarity algorithm and a distance comparison algorithm; the processing unit 902 is configured to use the feature comparison algorithm to compare the comprehensive feature information indicated by the fused feature vector with each candidate feature information in the biometric feature dataset, and to obtain a feature matching result corresponding to each candidate feature information, specifically to:

[0146] Using a similarity algorithm, respectively, similarity calculations are performed on the comprehensive feature information indicated by the fused feature vector and each candidate feature information in the biometric feature data set to obtain a similarity comparison result corresponding to each candidate feature information; and

[0147] Using a distance comparison algorithm, a distance index is calculated for each of the first distances between the biometric feature corresponding to the fused feature vector and the acquisition device at the time of acquisition, and the second distance corresponding to each candidate feature information in the biometric feature dataset, to obtain a distance index result corresponding to each candidate feature information; the distance index result is used to indicate the distance difference between the first distance and the second distance; the second distance corresponding to the candidate feature information is the distance between the biometric feature corresponding to the candidate feature information and the acquisition device at the time of acquisition;

[0148] Based on the similarity comparison result and the distance index result corresponding to each candidate feature information, a feature matching result corresponding to each candidate feature information is generated.

[0149] In one implementation, when the feature matching result is a weighted value of the similarity comparison result and the distance index result, the feature matching result satisfies the matching condition including: the weighted value is greater than a weighted threshold;

[0150] When the feature matching result includes a similarity comparison result and a distance index result, the feature matching result satisfies the matching condition including: the similarity comparison result is greater than a similarity threshold, and the distance index result is greater than a distance difference threshold.

[0151] According to one embodiment of the present application, the various units in the identity recognition device shown in Figure 9 can be individually or completely combined into one or several other units to form, or one (or some) of the units can be further divided into multiple functionally smaller units to form, which can achieve the same operation without affecting the realization of the technical effects of the embodiments of the present application. The above-mentioned units are divided based on logical functions. In actual applications, the function of one unit can also be implemented by multiple units, or the functions of multiple units can be implemented by one unit. In other embodiments of the present application, the identity recognition device can also include other units. In actual applications, these functions can also be implemented with the assistance of other units and can be implemented by the collaboration of multiple units. According to another embodiment of the present application, the identity recognition device shown in Figure 9 can be constructed, and the identity recognition method of the embodiment of the present application can be implemented by running a computer program (including program code) that can execute the steps involved in the corresponding methods shown in Figures 2 and 7 on a general computing device such as a computer that includes processing elements and storage elements such as a central processing unit (CPU), a random access memory medium (RAM), and a read-only memory medium (ROM). The computer program can be recorded on, for example, a computer-readable recording medium, loaded into the above-mentioned computing device via the computer-readable recording medium, and executed therein.

[0152] In an embodiment of the present application, continuous collection of a user's biometrics is supported to obtain an action sequence of the biometrics; wherein the action sequence includes multiple biometric graphs obtained from the continuous collection of the biometrics, and the feature changes between the biometric features contained in any two adjacent biometric graphs in the multiple biometric graphs are continuous. In this way, feature extraction processing can be performed on the multiple biometric graphs in the action sequence to obtain a feature vector for each biometric graph; the feature vector of any biometric graph represents the sub-feature information of the biometric from the semantic dimension of the biometric. For example, if the biometric is a palm, and the palm in the biometric graph is tilted, then the feature vector in the biometric graph represents the palm print characteristics when the palm is tilted. Furthermore, the feature vectors of each biometric graph are fused to generate a fused feature vector, which represents the comprehensive feature information of the biometric from the semantic dimension of the action sequence. Finally, identity recognition is performed based on the fused feature information of the biometrics to obtain an identification result indicating the object to which the biometric belongs. It can be seen from the above scheme that the embodiment of the present application provides a new biometric collection and identification scheme, which can continuously collect multiple biometric feature maps of the user when performing the target action (such as palm swiping action / palm swiping behavior), and analyze the comprehensive and rich vector representation of the biometric features based on multiple biometric feature maps for identity identification; compared with identity identification based on the feature vector of a single biometric feature map, it effectively improves the accuracy and reliability of the biometric features used for identity identification, thereby improving the accuracy of identity identification.

[0153] Figure 10 shows a schematic diagram of the structure of a computer device provided by an exemplary embodiment of the present application. Referring to Figure 10, the computer device includes a processor 1001, a communication interface 1002 and a computer-readable storage medium 1003. The processor 1001, the communication interface 1002 and the computer-readable storage medium 1003 can be connected via a bus or other means. The communication interface 1002 is used to receive and send data. The computer-readable storage medium 1003 can be stored in the memory of the computer device, and the computer-readable storage medium 1003 is used to store a computer program. The computer program includes program instructions, and the processor 1001 is used to execute the program instructions stored in the computer-readable storage medium 1003. The processor 1001 (or CPU (Central Processing Unit)) is the computing core and control core of the computer device, which is suitable for implementing one or more instructions, and is specifically suitable for loading and executing one or more instructions to implement the corresponding method flow or corresponding function.

[0154] The embodiment of the present application also provides a computer-readable storage medium (Memory), which is a memory device in a computer device for storing programs and data. It is understandable that the computer-readable storage medium here can include both built-in storage media in the computer device and, of course, extended storage media supported by the computer device. The computer-readable storage medium provides a storage space that stores the processing system of the computer device. In addition, one or more instructions suitable for being loaded and executed by the processor 1001 are also stored in the storage space. These instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory, or a non-volatile memory (non-volatile memory), such as at least one disk memory; optionally, it can also be at least one computer-readable storage medium located away from the aforementioned processor.

[0155] In one embodiment, the computer-readable storage medium stores one or more instructions; the processor 1001 loads and executes the one or more instructions stored in the computer-readable storage medium to implement the corresponding steps in the above-mentioned identity recognition method embodiment; in a specific implementation, the one or more instructions in the computer-readable storage medium are loaded by the processor 1001 and execute the following steps:

[0156] An action sequence for acquiring biometric features; the action sequence includes a plurality of biometric feature graphs continuously collected for the biometric features; and feature changes between two adjacent biometric feature graphs in the plurality of biometric feature graphs are continuous;

[0157] Performing feature extraction processing on each of the plurality of biometric feature graphs to obtain a feature vector of each biometric feature graph; the feature vector of any biometric feature graph represents sub-feature information of the biometric feature from a semantic dimension of any biometric feature graph;

[0158] The feature vectors of each biometric feature map are fused to generate a fused feature vector; the fused feature vector represents the comprehensive feature information of the biometric feature from the semantic dimension of the action sequence;

[0159] The biometric feature is identified based on the fused feature vector to obtain an identification result, which indicates the object to which the biometric feature belongs.

[0160] In one implementation, one or more instructions in the computer-readable storage medium are loaded by the processor 1001 and, when executing the action sequence for obtaining biometric features, specifically perform the following steps:

[0161] Continuously collect the biometrics of the target action to be performed in the business scenario, and obtain a video stream of biometrics; the video stream includes multiple initial biometric maps collected continuously, and each initial biometric map includes all or part of the features of the biometrics.

[0162] Preprocess the multiple initial biometric maps included in the video stream to obtain an action sequence of the biometrics.

[0163] Among them, the biometric map is the palm map of the object performing the target action, the biometrics included in the biometric map are the palmprint features of the palm, and the business scenario is a long-distance palm-swiping scenario.

[0164] In one implementation, the video stream includes N initial biometric maps, where N is an integer greater than 1; when one or more instructions in the computer-readable storage medium are loaded by the processor 1001 and execute preprocessing the multiple initial biometric maps included in the video stream to obtain an action sequence of the biometrics, the following steps are specifically executed:

[0165] Judge whether the feature change between the biometrics included in the (i + 1)-th initial biometric map in the video stream and the biometrics included in the i-th initial biometric map in the video stream is continuous; 1 ≤ i < N, and i is an integer.

[0166] If the feature change between the biometrics included in the (i + 1)-th initial biometric map and the biometrics included in the i-th initial biometric map in the video stream is continuous, add the (i + 1)-th initial biometric map to the action sequence of the biometrics, and judge whether the (i + 1)-th initial biometric map is the last one among the N initial biometric maps included in the video stream.

[0167] If the (i + 1)-th initial biometric map is not the last one among the N initial biometric maps included in the video stream, continue to judge whether the feature change between the biometrics included in the (i + 2)-th initial biometric map and the biometrics included in the (i + 1)-th initial biometric map is continuous until a candidate biometric map in the video stream is detected.

[0168] Among them, the candidate biometric map includes: the last initial biometric map among the N initial biometric maps; or, among the N initial biometric maps, an initial biometric map whose feature change with the biometrics included in the adjacent previous initial biometric map is not continuous.

[0169] In one implementation, one or more instructions in the computer-readable storage medium are loaded by the processor 1001 and further execute the following steps:

[0170] If the feature changes between the biometric features included in the i+1th initial biometric feature map and the biometric features included in the i-th initial biometric feature map in the video stream are not continuous, then an action sequence of the biometric features is determined; the action sequence includes i initial biometric feature maps in the video stream, and the i initial biometric feature maps are: the initial biometric feature maps that are collected in the video stream and are located before the i+1th initial biometric feature map.

[0171] In one implementation, if a deep learning model is used to determine whether the feature changes between the biometric features included in the (i+1)th initial biometric feature map and the biometric features included in the (i)th initial biometric feature map are continuous, one or more instructions in the computer-readable storage medium are loaded by the processor 1001 and the following steps are further executed:

[0172] Add the i+1th initial biometric feature map to the interference dataset;

[0173] Among them, the interference data set includes: an initial biometric feature map in which the feature changes between the biometric features in the video stream and the biometric features contained in the previous adjacent initial biometric feature map are not continuous; the interference data set serves as interference data for the deep learning model during the model training process.

[0174] In one implementation, one or more instructions in the computer-readable storage medium are loaded by the processor 1001 and further perform the following steps:

[0175] Performing image quality detection on each biometric feature image in the action sequence to obtain a quality detection result of each biometric feature image;

[0176] Filter out biometric features whose quality inspection results do not meet quality requirements from the action sequence;

[0177] Performing image enhancement processing on the biometric image whose quality inspection result does not meet the quality requirements to obtain an image-enhanced biometric image;

[0178] The enhanced biometric image is used to update the action sequence of the biometrics to obtain an updated target action sequence; the target action sequence includes: the enhanced biometric image after image enhancement and the biometric image whose quality detection result meets the quality requirements.

[0179] In one implementation, a biometric image whose quality inspection result does not meet the quality requirements is represented as a target biometric image; one or more instructions in a computer-readable storage medium are loaded by the processor 1001 and, when performing image enhancement processing on the biometric image whose quality inspection result does not meet the quality requirements to obtain an enhanced biometric image after image enhancement, the processor 1001 specifically performs the following steps:

[0180] Obtain the image processing requirements of the target biometric image and select an image enhancement algorithm that matches the image processing requirements;

[0181] Using an image processing algorithm that matches the image processing requirements, the target biometric image is enhanced to obtain an enhanced biometric image after image enhancement processing;

[0182] Among them, if the image processing requirement is to enhance the high-frequency information in the enhanced biometric feature image, the image enhancement algorithms that match the image processing requirements include: interpolation algorithms and learning algorithms under super-resolution algorithms.

[0183] In one implementation, one or more instructions in the computer-readable storage medium are loaded by the processor 1001 and, when performing identity recognition based on the fused feature vector for the biometric feature and obtaining the recognition result, specifically perform the following steps:

[0184] Acquire a biometric feature dataset, where the biometric feature dataset includes candidate feature information of biometric features of different objects;

[0185] Using the feature comparison algorithm, the comprehensive feature information indicated by the fused feature vector is compared with each candidate feature information in the biometric dataset to obtain the feature matching result corresponding to each candidate feature information;

[0186] Generate recognition results based on the feature matching results corresponding to each candidate feature information;

[0187] The recognition result indicates that candidate feature information exists in the biometric feature data set whose feature matching result satisfies the matching condition, or the recognition result indicates that candidate feature information does not exist in the biometric feature data set whose feature matching result satisfies the matching condition.

[0188] In one implementation, the feature comparison algorithm includes a similarity algorithm and a distance comparison algorithm. One or more instructions in the computer-readable storage medium are loaded by the processor 1001 and, when executing the feature comparison algorithm, the processor 1001 compares the comprehensive feature information indicated by the fused feature vector with each candidate feature information in the biometric feature dataset to obtain a feature matching result corresponding to each candidate feature information. Specifically, the processor 1001 performs the following steps:

[0189] Using a similarity algorithm, respectively, similarity calculations are performed on the comprehensive feature information indicated by the fused feature vector and each candidate feature information in the biometric feature data set to obtain a similarity comparison result corresponding to each candidate feature information; and

[0190] Using a distance comparison algorithm, a distance index is calculated for each of the first distances between the biometric feature corresponding to the fused feature vector and the acquisition device at the time of acquisition, and the second distance corresponding to each candidate feature information in the biometric feature dataset, to obtain a distance index result corresponding to each candidate feature information; the distance index result is used to indicate the distance difference between the first distance and the second distance; the second distance corresponding to the candidate feature information is the distance between the biometric feature corresponding to the candidate feature information and the acquisition device at the time of acquisition;

[0191] Based on the similarity comparison result and the distance index result corresponding to each candidate feature information, a feature matching result corresponding to each candidate feature information is generated.

[0192] In one implementation, when the feature matching result is a weighted value of the similarity comparison result and the distance index result, the feature matching result satisfies the matching condition including: the weighted value is greater than a weighted threshold;

[0193] When the feature matching result includes a similarity comparison result and a distance index result, the feature matching result satisfies the matching condition including: the similarity comparison result is greater than a similarity threshold, and the distance index result is greater than a distance difference threshold.

[0194] Based on the same inventive concept, the principles and beneficial effects of solving the problems provided by the computer device in the embodiment of the present application are similar to the principles and beneficial effects of solving the problems provided by the identity recognition method in the method embodiment of the present application. Please refer to the principles and beneficial effects of the implementation of the method. For the sake of concise description, they will not be repeated here.

[0195] The present application also provides a computer program product or computer program, which includes computer instructions stored in a computer-readable storage medium. A processor of a computer device reads the computer instructions from the computer-readable storage medium and executes the computer instructions, causing the computer device to perform the above-described identity recognition method.

[0196] Those skilled in the art will appreciate that the units and algorithmic steps of each example described in conjunction with the embodiments disclosed in this application can be implemented in electronic hardware, or a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution. Professional technical personnel may use different methods to implement the described functions for each specific application, but such implementation should not be considered beyond the scope of this application.

[0197] In the above embodiments, it can be implemented in whole or in part by software, hardware, firmware or any combination thereof. When software is used for implementation, it can be implemented in whole or in part in the form of a computer program product. The computer program product includes one or more computer instructions. When the computer program instructions are loaded and executed on a computer, the process or function according to the embodiment of the present application is generated in whole or in part. The computer can be a general-purpose computer, a special-purpose computer, a computer network, or other programmable device. The computer instructions can be stored in a computer-readable storage medium or transmitted via a computer-readable storage medium. The computer instructions can be transmitted from one website, computer, server or data center to another website, computer, server or data center via a wired (e.g., coaxial cable, optical fiber, digital line (DSL)) or wireless (e.g., infrared, wireless, microwave, etc.) method. The computer-readable storage medium can be any available medium that a computer can access or a data processing device such as a server or data center that includes one or more available media integrations. The available medium can be a magnetic medium (e.g., a floppy disk, a hard disk, a tape), an optical medium (e.g., a DVD), or a semiconductor medium (e.g., a solid-state drive (SSD)).

[0198] The above description is merely a specific embodiment of the present application, but the scope of protection of the present application is not limited thereto. Any technical object of a person skilled in the art that can be easily conceived of within the technical scope disclosed in this application should be included in the scope of protection of the present application. Therefore, the scope of protection of the present application should be based on the scope of protection of the claims.

Claims

1. An identity recognition method, characterized in that: The method is executed by a computer device, and the method includes: Obtaining an action sequence of a biometric feature; the action sequence includes a plurality of biometric feature maps continuously collected for the biometric feature; the feature changes between adjacent two of the plurality of biometric feature maps are continuous; Performing feature extraction processing on each of the plurality of biometric feature maps to obtain a feature vector of each biometric feature map; the feature vector of any biometric feature map represents sub-feature information of the biometric feature from the semantic dimension of any biometric feature map; Fusing the feature vectors of each biometric feature map to generate a fused feature vector; the fused feature vector represents comprehensive feature information of the biometric feature from the semantic dimension of the action sequence; Performing identity recognition on the biometric feature based on the fused feature vector to obtain a recognition result, where the recognition result indicates the object to which the biometric feature belongs.

2. The method according to claim 1, characterized in that The obtaining of the action sequence of the biometric feature includes: Continuously collecting the biometric feature of the target action to be performed in a business scenario to obtain a video stream of the biometric feature; the video stream includes a plurality of initial biometric feature maps continuously collected, and each initial biometric feature map includes all or part of the features of the biometric feature; Preprocessing the plurality of initial biometric feature maps included in the video stream to obtain the action sequence of the biometric feature; Wherein, the biometric feature map is a palm map of the palm of the object performing the target action, the biometric feature included in the biometric feature map is the palmprint feature of the palm, and the business scenario is a long-distance palm-sweeping scenario.

3. The method according to claim 1 or 2, characterized in that The video stream includes N initial biometric feature maps, where N is an integer greater than 1; the preprocessing of the plurality of initial biometric feature maps included in the video stream to obtain the action sequence of the biometric feature includes: Judging whether the feature change between the biometric feature included in the (i + 1)-th initial biometric feature map in the video stream and the biometric feature included in the i-th initial biometric feature map in the video stream is continuous; 1 ≤ i < N, and i is an integer; If the feature change between the biometric feature included in the (i + 1)-th initial biometric feature map and the biometric feature included in the i-th initial biometric feature map in the video stream is continuous, adding the (i + 1)-th initial biometric feature map to the action sequence of the biometric feature, and judging whether the (i + 1)-th initial biometric feature map is the last one of the N initial biometric feature maps included in the video stream; If the (i + 1)-th initial biometric feature map is not the last one of the N initial biometric feature maps included in the video stream, continue to judge whether the feature change between the biometric feature included in the (i + 2)-th initial biometric feature map and the biometric feature included in the (i + 1)-th initial biometric feature map is continuous until a candidate biometric feature map in the video stream is detected; The candidate biometric feature map includes: the last one of the N initial biometric feature maps; or an initial biometric feature map among the N initial biometric feature maps, in which the feature changes between the biometric features and the biometric features included in the adjacent previous initial biometric feature map are not continuous.

4. The method according to any one of claims 1 to 3, characterized in that: The method further comprises: If the feature changes between the biometric features included in the i+1th initial biometric feature map and the biometric features included in the i-th initial biometric feature map in the video stream are not continuous, then an action sequence of the biometric features is determined; the action sequence includes i-th initial biometric feature maps in the video stream, and the i-th initial biometric feature maps are: the initial biometric feature maps whose acquisition order in the video stream is before the i+1th initial biometric feature map.

5. The method according to any one of claims 1 to 4, characterized in that: If a deep learning model is used to implement the judgment of whether the feature changes between the biometric features included in the i+1th initial biometric feature graph and the biometric features included in the ith initial biometric feature graph are continuous, the method further includes: Adding the i+1th initial biometric feature map to the interference data set; Among them, the interference data set includes: an initial biometric feature map in which the feature changes between the biometric features in the video stream and the biometric features contained in the previous adjacent initial biometric feature map are not continuous; the interference data set serves as interference data for the deep learning model during the model training process.

6. The method according to any one of claims 1 to 5, characterized in that: The method further comprises: Performing image quality detection on each of the biometric characteristic images in the action sequence to obtain a quality detection result of each of the biometric characteristic images; Screening out biometric feature graphs whose quality inspection results do not meet quality requirements from the action sequence; Performing image enhancement processing on the biometric image whose quality inspection result does not meet the quality requirements to obtain an enhanced biometric image after image enhancement; The enhanced biometric image is used to update the action sequence of the biometric to obtain an updated target action sequence; the target action sequence includes: the enhanced biometric image after image enhancement and the biometric image whose quality detection result meets the quality requirements.

7. The method according to any one of claims 1 to 6, characterized in that: The biometric image whose quality inspection result does not meet the quality requirements is represented as a target biometric image; the image enhancement processing is performed on the biometric image whose quality inspection result does not meet the quality requirements to obtain an enhanced biometric image after image enhancement, including: Obtaining image processing requirements of the target biometric image, and selecting an image enhancement algorithm that matches the image processing requirements; Using the image processing algorithm, the target biological characteristic image is subjected to image enhancement processing to obtain an enhanced biological characteristic image after image enhancement processing; Among them, if the image processing requirement is to enhance the high-frequency information in the enhanced biometric feature map, the image enhancement algorithms that match the image processing requirement include: an interpolation algorithm and a learning algorithm under a super-resolution algorithm.

8. The method according to any one of claims 1 to 7, characterized in that: The performing identity recognition on the biometric feature based on the fused feature vector to obtain a recognition result includes: Acquire a biometric feature data set, wherein the biometric feature data set includes candidate feature information of biometric features of different objects; Using a feature comparison algorithm, respectively compare the comprehensive feature information indicated by the fused feature vector with each of the candidate feature information in the biometric feature data set to obtain a feature matching result corresponding to each of the candidate feature information; Generate a recognition result based on the feature matching result corresponding to each of the candidate feature information; The recognition result indicates that there is candidate feature information in the biometric data set whose feature matching result satisfies the matching condition, or the recognition result indicates that there is no candidate feature information in the biometric data set whose feature matching result satisfies the matching condition.

9. The method according to any one of claims 1 to 8, characterized in that The feature comparison algorithm includes a similarity algorithm and a distance comparison algorithm; the feature comparison algorithm is used to compare the comprehensive feature information indicated by the fused feature vector with each of the candidate feature information in the biometric feature data set, respectively, to obtain a feature matching result corresponding to each of the candidate feature information, including: Using the similarity algorithm, respectively, calculating the similarity of the comprehensive feature information indicated by the fused feature vector and each of the candidate feature information in the biometric feature data set, to obtain a similarity comparison result corresponding to each of the candidate feature information; and, By using the distance comparison algorithm, the first distance between the biological feature corresponding to the fused feature vector and the collection device when it is collected is respectively compared. distance, and performing distance index calculation on the second distance corresponding to each of the candidate feature information in the biometric feature data set to obtain a distance index result corresponding to each of the candidate feature information; the distance index result is used to indicate the distance difference between the first distance and the second distance; the second distance corresponding to the candidate feature information refers to the distance between the biometric feature corresponding to the candidate feature information and the collection device when it is collected; Based on the similarity comparison result and the distance index result corresponding to each of the candidate feature information, a feature matching result corresponding to each of the candidate feature information is generated.

10. The method according to any one of claims 1 to 9, characterized in that: When the feature matching result is a weighted value of the similarity comparison result and the distance index result, the feature matching result satisfies the matching condition including: the weighted value is greater than a weighted threshold; When the feature matching result includes the similarity comparison result and the distance index result, the feature matching result satisfies the matching condition including: the similarity comparison result is greater than a similarity threshold, and the distance index result is greater than a distance difference threshold.

11. An identity recognition device, characterized in that: The identity recognition device is mounted on a computer device, and the identity recognition device includes: An acquisition unit, configured to acquire an action sequence of a biometric feature; the action sequence includes a plurality of biometric feature graphs continuously collected for the biometric feature; and the feature changes between the biometric features included in two adjacent biometric feature graphs among the plurality of biometric feature graphs are continuous; A processing unit, configured to perform feature extraction processing on each of the plurality of biometric feature graphs to obtain a feature vector of each of the biometric feature graphs; the feature vector of any of the biometric feature graphs represents sub-feature information of the biometric feature from a semantic dimension of any of the biometric feature graphs; The processing unit is further used to fuse the feature vectors of each of the biometric feature images to generate a fused feature vector; the fused feature vector represents the comprehensive feature information of the biometric feature from the semantic dimension of the action sequence; The processing unit is further used to perform identity recognition on the biometric feature based on the fused feature vector to obtain a recognition result, wherein the recognition result indicates the object to which the biometric feature belongs.

12. A computer device, characterized in that: a processor adapted to execute a computer program; A computer-readable storage medium, wherein a computer program is stored in the computer-readable storage medium, and when the computer program is executed by the processor, the identity recognition method according to any one of claims 1 to 10 is implemented.

13. A computer-readable storage medium, characterized in that: The computer-readable storage medium stores a computer program, and the computer program is suitable for being loaded by a processor and executing the identity recognition method according to any one of claims 1 to 10.

14. A computer program product, characterized in that The computer program product comprises a computer program, and when the computer program is executed by a processor, the identity recognition method according to any one of claims 1 to 10 is implemented.

Images