A dynamic updating face recognition method, device and electronic equipment

Abstract

The application discloses a dynamic updating face recognition method and device and electronic equipment, the method comprises the following steps: acquiring system configuration parameters, the system configuration parameters include face recognition algorithm type identifier; according to the face recognition algorithm type identifier, the corresponding face recognition algorithm is called, the feature of the face picture to be detected is extracted, and the face feature vector is obtained; the data linked list corresponding to the face recognition algorithm type identifier is called from the database, the data linked list stores different face image identifiers and the face feature values obtained by the face recognition algorithm corresponding to the face recognition algorithm type identifier and different face image feature extraction; the face feature values in the data linked list are traversed and compared with the face feature vector respectively, the similarity is calculated, the face image with the maximum similarity and the corresponding personnel information are output; the application is convenient for updating and upgrading the face recognition algorithm, and can adapt to the rapid development and replacement of the face recognition technology.

Description

Technical Field

[0001] This application relates to the field of facial recognition technology, and more specifically, to a dynamically updated facial recognition method, apparatus, and electronic device. Background Technology

[0002] In recent years, with the rapid development of computer vision, big data, artificial intelligence, and machine learning technologies, facial recognition technology has experienced explosive growth in various countries, bringing great convenience to people's work and lives. Facial recognition technology is a biometric identification technology based on facial feature information. It uses cameras or webcams to capture images or video streams containing faces, automatically detects and tracks faces in the images, and then identifies and judges the detected faces. A facial recognition system mainly consists of four components: facial image acquisition and detection, facial image preprocessing, facial image feature extraction, and facial image matching and recognition.

[0003] Facial image feature extraction and facial image matching and recognition are key aspects of facial recognition technology applications. Facial image feature extraction involves extracting a string of numbers representing facial information. Facial image matching and recognition involves searching and matching the extracted facial image feature data with facial feature data templates stored in a database and outputting the matching result. This requires the use of facial recognition algorithms, which are the foundation of facial recognition technology. There are various facial recognition algorithms, and as research progresses, more algorithms will be applied in the market. Therefore, facial recognition systems need to be compatible with various facial recognition algorithms, enabling rapid updates and application in real-world products.

[0004] However, current facial recognition systems on the market all use a specific facial recognition algorithm. Facial feature extraction and comparative analysis are based on a single facial model, without considering the compatibility of facial recognition algorithms. This makes it inconvenient to update and upgrade facial recognition algorithms and cannot adapt to the rapid development and replacement of facial recognition technology. Summary of the Invention

[0005] In response to at least one defect or improvement need in the prior art, the present invention provides a dynamically updated face recognition method, device and electronic device, the purpose of which is to solve the problems of the existing face recognition system having a single face recognition method and inconvenient updates.

[0006] To achieve the above objectives, according to one aspect of the present invention, a dynamically updated face recognition method is provided, the method comprising the following steps:

[0007] Obtain system configuration parameters, which include a face recognition algorithm type identifier;

[0008] The corresponding face recognition algorithm is invoked according to the face recognition algorithm type identifier, and the face image to be detected is used to extract features to obtain a face feature vector;

[0009] Retrieve the data list corresponding to the face recognition algorithm type identifier from the database. The data list stores different face images and facial feature values ​​obtained by extracting features from different face images based on the face recognition algorithm corresponding to the face recognition algorithm type identifier.

[0010] Traverse the facial feature values ​​in the data linked list and compare them with the facial feature vectors respectively, calculate the similarity, and output the face image with the highest similarity and its corresponding personnel information.

[0011] Preferably, the above-mentioned dynamically updated face recognition method further includes:

[0012] Obtain an updated face recognition algorithm, wherein the updated face recognition algorithm has a corresponding face recognition algorithm type identifier;

[0013] All face images are extracted from the database, and features are re-extracted from all face images based on the updated face recognition algorithm to obtain updated facial feature values;

[0014] The updated face recognition algorithm type identifier, face image identifier, and corresponding facial feature update value are associated and stored in the database.

[0015] Preferably, in the above-mentioned dynamically updated face recognition method, the system configuration parameters also include the calculation method corresponding to the face recognition algorithm, the face feature value type, and the length parameter.

[0016] Preferably, in the above-mentioned dynamically updated face recognition method, different face recognition algorithms are configured as dynamic link libraries in the form of plug-ins;

[0017] The corresponding face recognition algorithm is retrieved from the dynamic link library using the face recognition algorithm type identifier as an index, and features are extracted from the face image to be detected; or,

[0018] The updated face recognition algorithm is retrieved from the dynamic link library using the face recognition algorithm type identifier as an index, and features are re-extracted from all face images extracted from the database.

[0019] According to a second aspect of the present invention, a dynamically updated face recognition device is also provided, comprising:

[0020] The configuration module is used to obtain system configuration parameters, which include a face recognition algorithm type identifier.

[0021] The face data extraction module is used to retrieve the corresponding face recognition algorithm according to the face recognition algorithm type identifier, extract features from the face image to be detected, and obtain a face feature vector.

[0022] The face data comparison module is used to retrieve the data linked list corresponding to the face recognition algorithm type identifier from the database. The data linked list stores different face image identifiers and facial feature values ​​obtained by extracting features from different face images based on the face recognition algorithm corresponding to the face recognition algorithm type identifier. The module also iterates through the facial feature values ​​in the data linked list and compares them with the face feature vectors respectively to calculate the similarity.

[0023] The results output module is used to obtain the calculation results of the face data comparison module and output the face image with the highest similarity and its corresponding personnel information.

[0024] Preferably, the aforementioned dynamically updated face recognition device further includes a feature re-extraction module;

[0025] The feature re-extraction module is used to obtain an updated face recognition algorithm, which has a corresponding face recognition algorithm type identifier; extract all face images from the database, and perform feature re-extraction on all face images based on the updated face recognition algorithm to obtain updated facial feature values;

[0026] It also associates the updated face recognition algorithm with the face recognition algorithm type identifier, face image identifier and its corresponding facial feature update value, and stores them in the database.

[0027] Preferably, the system configuration parameters of the aforementioned dynamically updated face recognition device also include the calculation method corresponding to the face recognition algorithm, the face feature value type, and the length parameter.

[0028] Preferably, in the aforementioned dynamically updated face recognition device, different face recognition algorithms are configured as dynamic link libraries in the form of plug-ins.

[0029] Preferably, in the above-mentioned dynamically updated face recognition device, the face data extraction module retrieves the corresponding face recognition algorithm from the dynamic link library using the face recognition algorithm type identifier as an index, and performs feature extraction on the face image to be detected;

[0030] The feature re-extraction module uses the face recognition algorithm type identifier as an index to obtain the updated face recognition algorithm from the dynamic link library, and performs feature re-extraction on all face images extracted from the database.

[0031] According to a third aspect of the invention, an electronic device is also provided, comprising at least one processing unit and at least one storage unit, wherein the storage unit stores a computer program that, when executed by the processing unit, causes the processing unit to perform the steps of any of the methods described above.

[0032] In summary, compared with the prior art, the above-described technical solutions conceived by this invention can achieve the following beneficial effects:

[0033] The dynamically updated face recognition method, device, and electronic device provided by this invention are compatible with various face recognition algorithms. The face data extraction module and face data comparison module can obtain information about the face recognition algorithm currently used by the system through a configuration module. By identifying different face recognition algorithm types and calculation methods, the corresponding face recognition algorithm can be invoked to perform the corresponding face feature value extraction and comparison process, thereby obtaining the compared face similarity results and achieving the purpose of dynamically controlling the face recognition algorithm. The feature re-extraction module assists in updating face feature value data when the face recognition algorithm is updated, facilitating direct comparison with the new face feature value data in the database when a new face recognition algorithm is enabled in the face recognition system, ensuring the consistency of the compared data. Attached Figure Description

[0034] To more clearly illustrate the technical solutions in the embodiments of this application, the accompanying drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of this application. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0035] Figure 1 This is a logic block diagram of a dynamically updated face recognition device provided in this embodiment;

[0036] Figure 2 This is a schematic diagram of the logical structure of the feature re-extraction module involved in this embodiment;

[0037] Figure 3 This is a flowchart illustrating the startup process of the face recognition device involved in this embodiment;

[0038] Figure 4 This is a flowchart illustrating the face recognition process involved in this embodiment;

[0039] Figure 5 This is a flowchart illustrating the feature extraction involved in this embodiment;

[0040] Figure 6 This is a flowchart illustrating a dynamically updated face recognition method provided in this embodiment;

[0041] Figure 7 This is a logical block diagram of the computer device provided in this embodiment. Detailed Implementation

[0042] The technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings.

[0043] The terms "first," "second," "third," etc., in the specification, claims, and accompanying drawings of this application are used to distinguish different objects, not to describe a specific order. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover non-exclusive inclusion. For example, a process, method, system, product, or apparatus that includes a series of steps or units is not limited to the listed steps or units, but may optionally include steps or units not listed, or may optionally include other steps or units inherent to these processes, methods, products, or apparatuses.

[0044] Figure 1 This is a logical block diagram of a dynamically updated face recognition device provided in this embodiment. The device can be implemented in software and / or hardware and can be integrated into an electronic device; see also... Figure 1 The device includes a configuration module, a face data extraction module, a face data comparison module, a result output module, and a database.

[0045] The configuration module is used to obtain system configuration parameters, which include a face recognition algorithm type identifier.

[0046] In this embodiment, the system pre-builds various face recognition algorithms. These algorithms are configured as dynamic link libraries (DLLs) in the form of plugins, and each algorithm has a unique corresponding type identifier. The executable code for each algorithm resides in a DLL file, which contains one or more functions that have been compiled, linked, and stored separately from the processes that use them. Using dynamic link libraries makes it easier to apply updates to the modules corresponding to each face recognition algorithm without affecting other parts.

[0047] The configuration module is responsible for acquiring the system's configuration parameters, mainly including parameters such as the face recognition algorithm type, calculation method, face feature value type and length, as well as log information, statistical information, and heartbeat time. These configuration parameters are then provided to the face data extraction module and the face data comparison module, respectively. When the face recognition system starts, this configuration module reads the system configuration file and determines the face recognition algorithm to be used based on the face recognition algorithm configuration information in the configuration file.

[0048] The face data extraction module is used to retrieve the corresponding face recognition algorithm according to the face recognition algorithm type identifier, extract features from the face image to be detected, and obtain face feature vectors.

[0049] In this embodiment, the face data extraction module is responsible for receiving the face image data stream sent by the face image acquisition and extracting face feature value data from the image.

[0050] After determining the face recognition algorithm to be used based on the face recognition algorithm type identifier in the system configuration parameters, the face data extraction module uses the face recognition algorithm type identifier as an index to retrieve the corresponding face recognition algorithm from the dynamic link library, performs feature extraction on the face image to be detected, and obtains the face feature vector.

[0051] In this embodiment, a plugin directory is constructed to store face recognition algorithm information in dynamic link libraries. This plugin directory uses the face recognition algorithm type identifier as an index, which makes it easy for the face data extraction module to quickly find and call the functions of the required face recognition algorithm from the dynamic link libraries.

[0052] The face data comparison module is used to retrieve the data linked list corresponding to the face recognition algorithm type identifier from the database. The data linked list stores different face image identifiers and facial feature values ​​obtained by extracting features from different face images based on the face recognition algorithm corresponding to the face recognition algorithm type identifier. The module also iterates through the facial feature values ​​in the data linked list and compares them with the face feature vectors respectively to calculate the similarity.

[0053] In this embodiment, the pre-built database stores data linked lists associated with each face recognition algorithm. Each data linked list stores different face image identifiers and facial feature values ​​obtained by the face recognition algorithm based on a certain face recognition algorithm type identifier when extracting features from different face images.

[0054] After the face data comparison module obtains the face recognition algorithm type identifier sent by the configuration module, it searches the database for the data list corresponding to the matching face recognition algorithm based on the face recognition algorithm type identifier. Then, it traverses the face image identifiers in the data list and obtains the corresponding facial feature values, and compares them with the face feature vectors output by the face data extraction module to calculate the similarity of the faces.

[0055] The results output module is used to obtain the calculation results of the face data comparison module and output the face image with the highest similarity and its corresponding personnel information.

[0056] In this embodiment, the result output module is responsible for analyzing and processing the similarity results obtained by the face data comparison module, and statistically analyzing the most similar face images and their personnel information according to the needs of the upper-layer application.

[0057] Furthermore, the aforementioned dynamically updated face recognition device also includes a feature re-extraction module;

[0058] The feature re-extraction module is a functional module that assists in updating facial feature value data when the facial recognition algorithm is updated. It is used to obtain the updated facial recognition algorithm, which has a corresponding facial recognition algorithm type identifier; extract all facial images from the database; perform feature re-extraction on all facial images based on the updated facial recognition algorithm to obtain updated facial feature values; and associate the updated facial recognition algorithm type identifier, facial image identifier and corresponding updated facial feature values ​​with the updated facial recognition algorithm and store them in the database.

[0059] If changes to the facial recognition algorithm lead to changes in facial feature value data, the facial feature value data previously stored in the database, extracted using the old algorithm, cannot be used in the face comparison and recognition process of the new algorithm. Therefore, it is necessary to re-extract the feature values. Thus, before applying the new facial recognition algorithm in the facial data extraction and comparison modules, the feature re-extraction module needs to re-extract the feature values ​​from the facial images stored in the database to obtain new facial feature value data. This new data is then stored in the database, allowing for direct comparison with the new facial feature value data when the facial recognition system uses the new algorithm, ensuring the consistency of the comparison data.

[0060] The feature re-extraction module is responsible for constructing a new facial feature value structure before updating the face recognition algorithm. It re-extracts facial feature value data from the face images stored in the database and then stores the new feature value data in the corresponding location in the database. In this embodiment, the feature re-extraction module uses the face recognition algorithm type identifier as an index to obtain the updated face recognition algorithm from the dynamic link library and performs feature re-extraction on all face images extracted from the database.

[0061] Figure 2 This is a schematic diagram of the logical structure of the feature re-extraction module involved in this invention; as shown... Figure 2 As shown, the feature re-extraction module comprises three parts: a database interface, a face image reading module, and a face feature value extraction module. The database interface provides query and update functions for face images and related data in the database. The face image reading module reads the face image data obtained from the database. The face feature value extraction module extracts face feature values ​​from the face image data and writes the new face feature value data to the corresponding location in the database through the database interface.

[0062] As an optional embodiment, the aforementioned dynamically updated face recognition device further includes a face data storage module; this face data storage module serves as an interaction bridge between the face data extraction module, the face data comparison module, and the database. Its main responsibilities include storing the face feature values ​​extracted from face images by the face data extraction module in the database and classifying and managing them; and searching for the data linked list corresponding to the matching face recognition algorithm in the database according to the face recognition algorithm type identifier sent by the configuration module and sending it to the face data comparison module.

[0063] In this solution, the system has various face recognition algorithms pre-built in. The face data extraction module and the face data comparison module can obtain the face recognition algorithm information currently used by the system through the configuration module. By identifying different face recognition algorithm types and calculation methods, the corresponding face recognition algorithm can be called to perform the corresponding face feature value extraction and comparison process, thereby obtaining the face similarity results of the comparison, and achieving the purpose of dynamically controlling the face recognition algorithm.

[0064] Figure 3 This is a flowchart illustrating the startup process of the face recognition device involved in this embodiment.

[0065] Step 301: The process begins.

[0066] Step 302: The configuration module reads the system configuration file to extract system configuration information, including face recognition algorithm information, face feature value information, log information, statistical information, and heartbeat time.

[0067] Step 303: The configuration module notifies the face data extraction module of the face recognition algorithm information.

[0068] Step 304: The face data extraction module determines the face feature value extraction algorithm for the face image.

[0069] Step 305: The configuration module notifies the face data storage module of the face feature value information.

[0070] Step 306: The face data storage module determines the storage structure and data information of face feature values.

[0071] Step 307: The configuration module notifies the face recognition algorithm information and face feature value information to the face data comparison module.

[0072] Step 308: The face data comparison module determines the data structure of face feature values ​​and its comparison and recognition algorithm.

[0073] Step 309: Start the heartbeat timer and statistics timer according to the configuration file, and the system will start up.

[0074] Figure 4 This is a flowchart illustrating the face recognition process involved in this embodiment.

[0075] Step 401: Begin the process.

[0076] Step 402: The face data extraction module receives the face image message sent by the image acquisition device.

[0077] Step 403: The face data extraction module parses the message to obtain the face image, extracts the face feature value data from the face image according to the face recognition algorithm, and then appends it to the face image message and sends it to the face data storage module.

[0078] Step 404: The face data storage module stores the received face image and its feature values ​​in the database, and then transmits the message containing the face feature values ​​to the face data comparison module.

[0079] Step 405: The face data comparison module parses the face feature value data in the message, uses the face recognition algorithm configured by the system to compare the face feature value data with the face feature value data stored in the database, calculates the face similarity, and notifies the result analysis output module of the result.

[0080] Step 406: The result analysis and output module outputs the results by statistically analyzing the most similar face images and their personnel information based on the requirements of the upper-layer application.

[0081] Step 407 completes the process.

[0082] Figure 5 This is a flowchart of the feature extraction involved in this embodiment.

[0083] Step 501 begins.

[0084] Step 502 involves iterating through the database interface to retrieve the face image data that needs to be updated with facial feature values, and then sending it to the face image reading module.

[0085] Step 503: The face image reading module is responsible for reading face image data and transmitting it to the face feature value extraction module.

[0086] Step 504: Extract facial feature value data from the face image using a new face recognition algorithm, and store the new facial feature value data into the database through the database interface at the location of the facial feature value corresponding to the face image.

[0087] Step 505 ends the process.

[0088] Figure 6 This is a flowchart illustrating a dynamically updated face recognition method provided in this embodiment, as shown below. Figure 6 As shown, the method includes the following steps:

[0089] Step 601: Obtain system configuration parameters, which include the face recognition algorithm type identifier, as well as the calculation method, face feature value type, and length parameters corresponding to the face recognition algorithm.

[0090] In this example, the system has various face recognition algorithms pre-built in, and these different face recognition algorithms are configured as dynamic link libraries in the form of plugins.

[0091] Step 602: Based on the face recognition algorithm type identifier, retrieve the corresponding face recognition algorithm, extract features from the face image to be detected, and obtain the face feature vector.

[0092] The corresponding face recognition algorithm is retrieved from the dynamic link library using the face recognition algorithm type identifier as an index, and features are extracted from the face image to be detected.

[0093] Step 603 retrieves the data list corresponding to the face recognition algorithm type identifier from the database. The data list stores different face image identifiers and facial feature values ​​obtained by extracting features from different face images based on the face recognition algorithm corresponding to the face recognition algorithm type identifier.

[0094] Step 604: Traverse the facial feature values ​​in the data linked list and compare them with the facial feature vectors respectively, calculate the similarity, and output the face image with the highest similarity and its corresponding personnel information.

[0095] As a preferred embodiment, the above-described dynamically updated face recognition method further includes a feature re-extraction step, specifically comprising:

[0096] S1 obtains an updated face recognition algorithm, wherein the updated face recognition algorithm has a corresponding face recognition algorithm type identifier;

[0097] S2 extracts all face images from the database, and performs feature re-extraction on all face images based on the updated face recognition algorithm to obtain updated facial feature values;

[0098] In this example, the updated face recognition algorithm is retrieved from the dynamic link library using the face recognition algorithm type identifier as an index, and features are re-extracted from all face images extracted from the database.

[0099] S3 associates the updated face recognition algorithm with the face recognition algorithm type identifier, face image identifier, and corresponding facial feature update value, and stores them in the database.

[0100] It should be noted that although the operations of the methods of the embodiments of this specification are described in a specific order in the above embodiments, this does not require or imply that these operations must be performed in that specific order, or that all the operations shown must be performed to achieve the desired result. On the contrary, the steps depicted in the flowcharts may be executed in a different order. Additionally or alternatively, certain steps may be omitted, multiple steps may be combined into one step, and / or one step may be broken down into multiple steps.

[0101] This embodiment also provides an electronic device, such as... Figure 7 As shown, it includes at least one processor and at least one memory, wherein the memory stores a computer program. When the computer program is executed by the processor, it causes the processor to perform the steps of the dynamically updated face recognition method described above. The specific steps are not described in detail here. In this embodiment, the types of processor and memory are not specifically limited. For example, the processor can be a microprocessor, a digital information processor, an on-chip programmable logic system, etc.; the memory can be volatile memory, non-volatile memory, or a combination thereof.

[0102] The electronic device can also communicate with one or more external devices (such as a keyboard, pointing terminal, display, etc.), one or more terminals that enable users to interact with the electronic device, and / or any terminal that enables the electronic device to communicate with one or more other computing terminals (such as a network card, modem, etc.). This communication can be performed via an input / output (I / O) interface. Furthermore, the electronic device can also communicate with one or more networks (such as a Local Area Network (LAN), a Wide Area Network (WAN), and / or a public network, such as the Internet) via a network adapter.

[0103] This embodiment also provides a computer-readable medium storing a computer program executable by an electronic device. When the computer program is run on the electronic device, it causes the electronic device to perform the steps of the dynamically updated face recognition method described above. Types of computer-readable media include, but are not limited to, storage media such as SD cards, USB flash drives, fixed hard drives, and external hard drives.

[0104] Those skilled in the art will readily understand that the above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of the present invention should be included within the scope of protection of the present invention.

Claims

1. A dynamically updated face recognition method, characterized in that, include: Obtain system configuration parameters, which include a face recognition algorithm type identifier; different face recognition algorithms are configured as dynamic link libraries in the form of plugins; Each face recognition algorithm has a unique corresponding face recognition algorithm type identifier; a plugin directory is constructed to store face recognition algorithms in dynamic link libraries, and the plugin directory is indexed by the face recognition algorithm type identifier; the executable code of different face recognition algorithms is located in a DLL file, which contains one or more functions that have been compiled, linked and stored separately from the processes that use them; The corresponding face recognition algorithm is invoked according to the face recognition algorithm type identifier, and the face image to be detected is used to extract features to obtain a face feature vector; Retrieve the data list corresponding to the face recognition algorithm type identifier from the database. The data list stores different face image identifiers and facial feature values ​​obtained by extracting features from different face images based on the face recognition algorithm corresponding to the face recognition algorithm type identifier. When the face recognition algorithm is updated, the system assists in updating facial feature value data. It associates the updated face recognition algorithm type identifier, face image identifier, and their corresponding updated facial feature values ​​with the updated algorithm and stores them in the database. Before updating the face recognition algorithm, a new facial feature value structure is constructed, and facial feature value data is re-extracted from the face images stored in the database. The new feature value data is then stored in the corresponding location in the database. Using the face recognition algorithm type identifier as an index, the updated face recognition algorithm is retrieved from the dynamic link library, and features are re-extracted from all face images extracted from the database. Traverse the facial feature values ​​in the data linked list and compare them with the facial feature vectors respectively, calculate the similarity, and output the face image with the highest similarity and its corresponding personnel information.

2. The dynamically updated face recognition method as described in claim 1, characterized in that, Also includes: Obtain an updated face recognition algorithm, wherein the updated face recognition algorithm has a corresponding face recognition algorithm type identifier; All face images are extracted from the database, and features are re-extracted from all face images based on the updated face recognition algorithm to obtain updated facial feature values; The updated face recognition algorithm type identifier, face image identifier, and corresponding facial feature update value are associated and stored in the database.

3. The dynamically updated face recognition method as described in claim 2, characterized in that, The system configuration parameters also include the calculation method, facial feature value type, and length parameter corresponding to the facial recognition algorithm.

4. The dynamically updated face recognition method as described in claim 1 or 3, characterized in that, The corresponding face recognition algorithm is retrieved from the dynamic link library using the face recognition algorithm type identifier as an index, and features are extracted from the face image to be detected; or, The updated face recognition algorithm is retrieved from the dynamic link library using the face recognition algorithm type identifier as an index, and features are re-extracted from all face images extracted from the database.

5. A dynamically updated face recognition device, characterized in that, include: A configuration module is used to obtain system configuration parameters, including a face recognition algorithm type identifier. Different face recognition algorithms are configured as dynamic link libraries in the form of plugins. Each face recognition algorithm has a unique corresponding face recognition algorithm type identifier. A plugin directory is constructed to store the face recognition algorithms in the dynamic link libraries, and the plugin directory is indexed by the face recognition algorithm type identifier. The executable code of different face recognition algorithms is located in a DLL file, which contains one or more functions that have been compiled, linked, and stored separately from the processes that use them. The face data extraction module is used to retrieve the corresponding face recognition algorithm according to the face recognition algorithm type identifier, extract features from the face image to be detected, and obtain a face feature vector. The face data comparison module is used to retrieve the data linked list corresponding to the face recognition algorithm type identifier from the database. The data linked list stores different face image identifiers and facial feature values ​​obtained by extracting features from different face images based on the face recognition algorithm corresponding to the face recognition algorithm type identifier. The module also iterates through the facial feature values ​​in the data linked list and compares them with the face feature vectors respectively to calculate the similarity. The feature re-extraction module is a functional module that assists in updating facial feature value data when the face recognition algorithm is updated. It is used to associate the face recognition algorithm type identifier, face image identifier, and their corresponding updated facial feature values ​​with the updated face recognition algorithm and store them in the database. Before updating the face recognition algorithm, it constructs a new facial feature value structure, re-extracts facial feature value data from the face images stored in the database, and then stores the new feature value data in the corresponding location in the database. The feature re-extraction module uses the face recognition algorithm type identifier as an index to retrieve the updated face recognition algorithm from the dynamic link library and performs feature re-extraction on all face images extracted from the database. The results output module is used to obtain the calculation results of the face data comparison module and output the face image with the highest similarity and its corresponding personnel information.

6. The dynamically updated face recognition device as described in claim 5, characterized in that, It also includes a feature re-extraction module; The feature re-extraction module is used to obtain an updated face recognition algorithm, which has a corresponding face recognition algorithm type identifier; extract all face images from the database, and perform feature re-extraction on all face images based on the updated face recognition algorithm to obtain updated facial feature values; It also associates the updated face recognition algorithm with the face recognition algorithm type identifier, face image identifier and its corresponding facial feature update value, and stores them in the database.

7. The dynamically updated face recognition device as described in claim 6, characterized in that, The system configuration parameters also include the calculation method, facial feature value type, and length parameter corresponding to the facial recognition algorithm.

8. The dynamically updated face recognition device as described in claim 6, characterized in that, The face data extraction module uses the face recognition algorithm type identifier as an index to retrieve the corresponding face recognition algorithm from the dynamic link library and extracts features from the face image to be detected. The feature re-extraction module uses the face recognition algorithm type identifier as an index to obtain the updated face recognition algorithm from the dynamic link library, and performs feature re-extraction on all face images extracted from the database.

9. An electronic device, characterized in that, It includes at least one processing unit and at least one storage unit, wherein the storage unit stores a computer program that, when executed by the processing unit, causes the processing unit to perform the steps of the method according to any one of claims 1 to 4.

Images