Face recognition method and device, electronic equipment and readable storage medium

Abstract

The application provides a face recognition method and device, electronic equipment and readable storage medium, and relates to the technical field of face recognition. The method comprises the following steps: determining a candidate registered user face matched with a to-be-recognized face image based on a first basic face feature corresponding to the to-be-recognized face image and a second basic face feature corresponding to each registered user face in a face database; in the case that it is determined that the candidate registered user face has a similar registered user face, obtaining a first face feature similarity between the candidate registered user face and the to-be-recognized face image and a second face feature similarity between the similar registered user face and the to-be-recognized face image; and determining a target registered user face corresponding to the to-be-recognized face image from the candidate registered user face and the similar registered user face based on the first face feature similarity and the second face feature similarity, thereby solving the technical problems that face recognition cannot be better performed and misrecognition cannot be avoided in the prior art.

Description

Technical Field

[0001] This invention relates to the field of facial recognition technology, and more particularly to a facial recognition method, apparatus, electronic device, and readable storage medium. Background Technology

[0002] Currently, with the development of pattern recognition and image processing technologies, facial recognition technology is receiving increasing attention. It is commonly used in identity recognition systems, such as access control systems, attendance systems, bank self-service systems, and public security systems, to quickly identify individuals.

[0003] In existing technologies, feature values ​​are extracted from the face to be identified, and the matching degree between these feature values ​​and those in a database is obtained. The identity of the face to be identified is then determined based on the feature values ​​in the database that have a matching degree greater than a preset threshold. However, when multiple individuals with highly similar faces exist in the database, this method is prone to misidentification.

[0004] Therefore, how to better perform facial recognition and avoid false recognition is a technical problem that technical personnel in related fields urgently need to solve. Summary of the Invention

[0005] This invention provides a face recognition method, apparatus, electronic device, and readable storage medium to address the shortcomings of existing technologies in achieving better face recognition and avoiding false recognition, thereby improving the accuracy of face recognition results.

[0006] This invention provides a face recognition method, comprising:

[0007] Based on the first basic facial features corresponding to the face image to be identified and the second basic facial features corresponding to the faces of each registered user in the face database, candidate registered user faces that match the face image to be identified are determined.

[0008] If it is determined that the candidate registered user's face has a similar registered user's face, the first facial feature similarity between the candidate registered user's face and the face image to be identified, and the second facial feature similarity between the similar registered user's face and the face image to be identified are obtained.

[0009] Based on the numerical comparison results of the first facial feature similarity and the second facial feature similarity, the target registered user face corresponding to the face image to be identified is determined from the candidate registered user faces and the similar registered user faces.

[0010] According to a face recognition method provided by the present invention, the step of obtaining the first facial feature similarity between the candidate registered user's face and the face image to be recognized includes:

[0011] If the candidate registered user's face has additional facial features in the face database, the similarity of the first facial features is determined based on the first additional facial features corresponding to the face image to be identified and the second additional facial features corresponding to the candidate registered user's face.

[0012] If no additional facial features corresponding to the candidate registered user's face exist in the face database, the similarity of the first facial features is determined based on the first basic facial features corresponding to the face image to be identified and the second basic facial features corresponding to the candidate registered user's face. The feature point density of the additional facial features is greater than the feature point density of the basic facial features.

[0013] According to a face recognition method provided by the present invention, the step of obtaining the second facial feature similarity between the similar registered user's face and the face image to be identified includes:

[0014] If the face database contains additional face features corresponding to the similar registered user's face, the second face feature similarity is determined based on the first additional face feature corresponding to the face image to be identified and the third additional face feature corresponding to the similar registered user's face.

[0015] If no additional facial features corresponding to the similar registered user's face exist in the face database, the second facial feature similarity is determined based on the first basic facial features corresponding to the face image to be identified and the third basic facial features corresponding to the similar registered user's face. The feature point density of the additional facial features is greater than the feature point density of the basic facial features.

[0016] According to a face recognition method provided by the present invention, determining the target registered user face corresponding to the face image to be recognized from the candidate registered user faces and the similar registered user faces based on the numerical comparison result of the first face feature similarity and the second face feature similarity includes:

[0017] If the similarity of the first facial feature is greater than that of the second facial feature, the candidate registered user face is determined as the target registered user face corresponding to the face image to be identified;

[0018] If the similarity of the second facial feature is greater than that of the first facial feature, the similar registered user face is identified as the target registered user face corresponding to the face image to be identified.

[0019] According to a face recognition method provided by the present invention, the method further includes:

[0020] Obtain the first similarity between the fourth basic facial feature corresponding to the face image to be added to the database and the second basic facial feature corresponding to the face of a registered user in the face database;

[0021] If the first similarity is greater than a preset threshold, a similarity association relationship is constructed between the face image to be added to the database and the face of the registered user.

[0022] Extract the fourth additional facial feature corresponding to the face image to be added to the database, and store the fourth basic facial feature and the fourth additional facial feature corresponding to the face image to be added to the database.

[0023] According to a face recognition method provided by the present invention, the step of determining candidate registered user faces that match the face image to be recognized based on a first basic face feature corresponding to the face image to be recognized and a second basic face feature corresponding to each registered user face in the face database includes:

[0024] Obtain the second similarity between the first basic facial features corresponding to the face image to be identified and the second basic facial features corresponding to the face of the currently registered user;

[0025] If the second similarity is determined to be greater than a preset threshold, the face of the currently registered user is determined as the face of the candidate registered user;

[0026] If the second similarity is determined to be no greater than a preset threshold, the above steps are repeated based on the face of the next registered user until the face of the candidate registered user is determined.

[0027] The present invention also provides a face recognition device, comprising:

[0028] The face matching module is used to determine candidate registered user faces that match the face image to be identified based on the first basic face features corresponding to the face image to be identified and the second basic face features corresponding to the faces of each registered user in the face database.

[0029] The similarity comparison module is used to obtain, when it is determined that there is a similar registered user face to the candidate registered user face, the first facial feature similarity between the candidate registered user face and the face image to be identified, and the second facial feature similarity between the similar registered user face and the face image to be identified;

[0030] The face recognition module is used to determine the target registered user face corresponding to the face image to be recognized from the candidate registered user faces and the similar registered user faces based on the numerical comparison results of the first face feature similarity and the second face feature similarity.

[0031] The present invention also provides an electronic device, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the program to implement the face recognition method as described above.

[0032] The present invention also provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the face recognition method as described above.

[0033] The present invention also provides a computer program product, including a computer program that, when executed by a processor, implements the face recognition method as described above.

[0034] The face recognition method, apparatus, electronic device, and readable storage medium provided by this invention quickly determine candidate registered user faces that match the face image to be recognized by using a first basic face feature corresponding to the face image to be recognized and a second basic face feature corresponding to each registered user face in the face database. Furthermore, when it is determined that there are similar registered user faces among the candidate registered user faces, the target registered user face with a higher feature similarity to the face image to be recognized is selected from the candidate registered user faces and similar registered user faces. This avoids the problem of misidentification when there are multiple people with high facial similarity in the database, thereby improving the accuracy and reliability of face recognition results and solving the technical problems of existing technologies that cannot better perform face recognition and avoid misidentification. Attached Figure Description

[0035] To more clearly illustrate the technical solutions in this invention or the prior art, the drawings used in the description of the embodiments or the prior art will be briefly introduced one by one below. Obviously, the drawings described below are some embodiments of this invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0036] Figure 1 This is one of the flowcharts illustrating the face recognition method provided in this embodiment of the invention;

[0037] Figure 2 This is a second schematic flowchart of the face recognition method provided in this embodiment of the invention;

[0038] Figure 3 This is the third flowchart illustrating the face recognition method provided in this embodiment of the invention;

[0039] Figure 4 This is the fourth flowchart illustrating the face recognition method provided in this embodiment of the invention;

[0040] Figure 5This is the fifth flowchart illustrating the face recognition method provided in this embodiment of the invention;

[0041] Figure 6 This is the sixth flowchart illustrating the face recognition method provided in this embodiment of the invention;

[0042] Figure 7 This is a schematic diagram of the structure of the face recognition device provided in an embodiment of the present invention;

[0043] Figure 8 This is a schematic diagram of the structure of the electronic device provided in an embodiment of the present invention. Detailed Implementation

[0044] To make the objectives, technical solutions, and advantages of this invention clearer, the technical solutions of this invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some, not all, of the embodiments of this invention. All other embodiments obtained by those skilled in the art based on the embodiments of this invention without creative effort are within the scope of protection of this invention.

[0045] The following is combined Figures 1-6 This invention describes the face recognition method provided by the present invention. For example... Figure 1 As shown, the present invention provides a face recognition method, comprising:

[0046] Step 101: Based on the first basic facial features corresponding to the face image to be identified and the second basic facial features corresponding to the faces of each registered user in the face database, determine the candidate registered user faces that match the face image to be identified.

[0047] The first basic facial feature is obtained by feature extraction from the image of the face to be identified, and includes all basic facial features used to identify the face to be identified. The second basic facial feature is obtained by feature extraction from the image of the face corresponding to the registered user's face, and includes all basic facial features used to identify the registered user's face pair.

[0048] Furthermore, the number of candidate registered user faces is one. A candidate registered user face refers to a registered user face whose similarity to the first basic face feature corresponding to the face image to be identified is greater than a preset threshold.

[0049] Furthermore, the face database stores the face feature data corresponding to the faces of each registered user. The face feature data includes the second basic face features corresponding to the faces of the registered users. The face database also stores the face identification information of each registered user's face, which can be a name, number, or ID.

[0050] Step 102: If it is determined that there are similar registered user faces among the candidate registered user faces, obtain the first facial feature similarity between the candidate registered user face and the face image to be identified, and the second facial feature similarity between the similar registered user face and the face image to be identified.

[0051] In one embodiment, the face database stores binary variables of similar faces corresponding to each registered user's face, as well as similar face identifier information. The binary variables of similar faces take either a first variable value or a second variable value. The first variable value indicates that a registered user's face has a similar face to another registered user, while the second variable value indicates that a registered user's face does not have a similar face to another registered user. The similar face identifier information includes the face identifier information of the similar registered user faces corresponding to the registered user's face.

[0052] Furthermore, if the value of the binary variable of the similar face corresponding to the candidate registered user's face is determined to be the first variable value, the name, number or ID of the corresponding similar registered user's face can be found based on the similar face identifier information of the candidate registered user's face, thereby obtaining the facial feature data of the corresponding similar registered user's face.

[0053] Step 103: Based on the numerical comparison results of the first face feature similarity and the second face feature similarity, determine the target registered user face corresponding to the face image to be identified from the candidate registered user faces and similar registered user faces.

[0054] Here, the target registered user face refers to the registered user face among the candidate registered user faces and similar registered user faces that has a higher feature similarity to the face image to be identified. Furthermore, based on the registered user information of the target registered user face, the identity information of the face to be identified can be determined, thereby achieving the purpose of face recognition.

[0055] Steps 101 to 103 above quickly identify candidate registered user faces that match the face image to be identified by using the first basic face features corresponding to the face image to be identified and the second basic face features corresponding to the faces of each registered user in the face database. Thus, when it is determined that there are similar registered user faces among the candidate registered user faces, the target registered user face with higher feature similarity to the face image to be identified is selected from the candidate registered user faces and similar registered user faces. This avoids the problem of misidentification when there are multiple people with high facial similarity in the database, thereby improving the accuracy and reliability of face recognition results and solving the technical problems of existing technologies that cannot better perform face recognition and avoid misidentification.

[0056] Furthermore, since the face database pre-stores the correspondence between each registered user's face and its similar registered user faces, it is not necessary to traverse the entire face database to determine the candidate registered user face that matches the face image to be identified. It is only necessary to match the candidate registered user faces whose similarity is greater than a preset threshold. Then, by using the correspondence between the matched candidate registered user faces and their corresponding similar registered user faces, it is possible to quickly filter out the target registered user face with a higher feature similarity to the face image to be identified from the candidate registered user face and its corresponding similar registered user faces. This improves the accuracy of face recognition results, simplifies the face recognition process, and increases the efficiency of face recognition.

[0057] In one embodiment, such as Figure 2 As shown, step 102 above includes steps 201 to 202, wherein:

[0058] Step 201: If there are additional facial features corresponding to candidate registered user faces in the face database, determine the similarity of the first facial features based on the first additional facial features corresponding to the face image to be identified and the second additional facial features corresponding to the candidate registered user faces.

[0059] Furthermore, the first additional facial feature is obtained based on feature extraction from the image of the face to be identified, and includes all additional facial features used to identify the face to be identified. The feature point density of the first additional facial feature is greater than the feature point density of the first basic facial feature. The first additional facial feature also includes all or some of the basic facial features that identify the face to be identified.

[0060] Furthermore, the second additional facial feature is obtained by feature extraction from the facial image corresponding to the candidate registered user's face, and includes all additional facial features used to identify the candidate registered user's face pair. The feature point density of the second additional facial feature is greater than the feature point density of the second basic facial feature. The second additional facial feature also includes all or some of the basic facial features that identify the candidate registered user's face.

[0061] Step 202: In the absence of additional facial features corresponding to the candidate registered user's face in the face database, the similarity of the first facial features is determined based on the first basic facial features corresponding to the face image to be identified and the second basic facial features corresponding to the candidate registered user's face. The feature point density of the additional facial features is greater than the feature point density of the basic facial features.

[0062] Steps 201 to 202 above improve face matching efficiency by quickly identifying candidate registered user faces that match the face image to be identified using basic face features with low feature point density during the first face matching. In the second face matching, additional face features with higher feature point density are selected to accurately identify the target registered user face corresponding to the face image to be identified, thereby improving face recognition performance. By using face features with different feature point densities for recognition and matching at different face matching stages, and by using face features adapted to each face matching stage based on the matching task and matching characteristics of each face matching stage, face recognition performance is improved while face matching efficiency is increased.

[0063] In one embodiment, such as Figure 3 As shown, step 102 above includes steps 301 to 302, wherein:

[0064] Step 301: If there are additional facial features corresponding to similar registered user faces in the face database, determine the similarity of the second facial feature based on the first additional facial feature corresponding to the face image to be identified and the third additional facial feature corresponding to similar registered user faces.

[0065] Furthermore, the first additional facial feature is obtained based on feature extraction from the image of the face to be identified, and includes all additional facial features used to identify the face to be identified. The feature point density of the first additional facial feature is greater than the feature point density of the first basic facial feature. The first additional facial feature also includes all or some of the basic facial features that identify the face to be identified.

[0066] Furthermore, the third additional facial feature is obtained by feature extraction from the facial images corresponding to similar registered user faces, and includes all additional facial features used to identify similar registered user face pairs. The feature point density of the third additional facial feature is greater than that of the third basic facial feature. The third additional facial feature also includes all or some of the basic facial features that identify similar registered user faces.

[0067] Step 302: If there are no additional facial features corresponding to similar registered user faces in the face database, determine the similarity of the second facial features based on the first basic facial features corresponding to the face image to be identified and the third basic facial features corresponding to similar registered user faces. The feature point density of the additional facial features is greater than the feature point density of the basic facial features.

[0068] Steps 301 to 302 above improve face matching efficiency by quickly identifying candidate registered user faces that match the face image to be identified using basic face features with low feature point density during the first face matching. In the second face matching, additional face features with higher feature point density are selected to accurately identify the target registered user face corresponding to the face image to be identified, thereby improving face recognition performance. By using face features with different feature point densities for recognition and matching at different face matching stages, and by using face features adapted to each face matching stage based on the matching task and matching characteristics of each face matching stage, face recognition performance is improved while face matching efficiency is increased.

[0069] In one embodiment, such as Figure 4 As shown, step 103 above includes steps 401 to 402, wherein:

[0070] Step 401: If the similarity of the first facial feature is greater than that of the second facial feature, the candidate registered user's face is determined as the target registered user's face corresponding to the face image to be identified.

[0071] In one embodiment, the first facial feature similarity is calculated based on a first additional facial feature corresponding to the face image to be identified and a second additional facial feature corresponding to the candidate registered user's face. The second facial feature similarity is calculated based on the first additional facial feature corresponding to the face image to be identified and a third additional facial feature corresponding to a similar registered user's face.

[0072] In one embodiment, the first facial feature similarity is calculated based on the first basic facial features corresponding to the face image to be identified and the second basic facial features corresponding to the candidate registered user's face. The second facial feature similarity is calculated based on the first basic facial features corresponding to the face image to be identified and the third basic facial features corresponding to the similar registered user's face.

[0073] In one embodiment, the first facial feature similarity is calculated based on a first additional facial feature corresponding to the face image to be identified and a second additional facial feature corresponding to the candidate registered user's face. The second facial feature similarity is calculated based on a first basic facial feature corresponding to the face image to be identified and a third basic facial feature corresponding to a similar registered user's face.

[0074] In one embodiment, the first facial feature similarity is calculated based on a first basic facial feature corresponding to the face image to be identified and a second basic facial feature corresponding to the candidate registered user's face. The second facial feature similarity is calculated based on a first additional facial feature corresponding to the face image to be identified and a third additional facial feature corresponding to the similar registered user's face.

[0075] It should be noted that in the first two embodiments, the first facial feature similarity and the first facial feature similarity are both calculated based on the additional facial features or the basic facial features. Therefore, the two can be directly compared to determine the target registered user's face with a higher similarity to the face image to be identified.

[0076] In the latter two embodiments, one of the first facial feature similarities is calculated based on additional facial features, while the other is calculated based on basic facial features. Even so, the two can still be directly compared to determine the target registered user's face with a higher similarity to the face image to be identified, for the following reasons:

[0077] If the "registered user face with additional facial features" and the face to be identified belong to the same person, the facial feature similarity score calculated based on the additional facial features will be higher, thus further confirming that the "registered user face with additional facial features" is the target registered user face corresponding to the face image to be identified. The "registered user face with additional facial features" can be a candidate registered user face or a similar registered user face.

[0078] If the "registered user face with additional facial features" does not belong to the same person as the face to be identified, the facial feature similarity value calculated based on the additional facial features will be lower. This makes it more certain that the "registered user face with additional facial features" is not the target registered user face corresponding to the face image to be identified. Consequently, it can be determined that the "registered user face without additional facial features" is the target registered user face corresponding to the face image to be identified.

[0079] Step 402: If the similarity of the second face feature is greater than that of the first face feature, the similar registered user face is identified as the target registered user face corresponding to the face image to be identified.

[0080] In one embodiment, such as Figure 5 As shown, the face recognition method provided by the present invention further includes steps 501 to 503, wherein:

[0081] Step 501: Obtain the first similarity between the fourth basic facial feature corresponding to the face image to be added to the database and the second basic facial feature corresponding to the face of a registered user in the face database.

[0082] Step 502: If the first similarity is greater than a preset threshold, construct a similarity association between the face image to be added to the database and the face of the registered user.

[0083] Specifically, when the first similarity is greater than a preset threshold, the binary variables of similar faces corresponding to the face to be added to the database and the registered user's face are both set to the first variable value. The face identifier information of the registered user's face is stored as similar face identifier information in the data item corresponding to the face to be added to the database, and the face identifier information of the face to be added to the database is also stored as similar face identifier information in the data item corresponding to the registered user's face. This completes the process of constructing the similarity association between the face image to be added to the database and the registered user's face. The first variable value is used to indicate that the registered user's face has a similar registered user face.

[0084] Furthermore, if the first similarity is not greater than a preset threshold, the binary variable representing similar faces corresponding to the face to be added to the database is set as the second variable value. The second variable value is used to indicate that there are no similar faces to the registered user's face. If a new face to be added to the database is similar to the current face to be added to the database, the binary variable representing similar faces corresponding to the current face to be added to the database can be changed from the second variable value to the first variable value.

[0085] Step 503: Extract the fourth additional facial feature corresponding to the face image to be added to the database, and store the fourth basic facial feature and the fourth additional facial feature corresponding to the face image to be added to the face database.

[0086] It should be noted that, in order to save storage resources in the face database, only the facial feature data corresponding to the registered user's face is stored in the face database, and not the face image corresponding to the registered user's face. Therefore, when it is determined that there is a similar correspondence between the face image to be added to the database and a certain registered user's face, only the additional facial features corresponding to the face image to be added to the database can be extracted, but not the additional facial features corresponding to the registered user's face. However, as long as either the face image to be added to the database or the registered user's face has additional facial features, the two can be distinguished. Therefore, even if the face database only stores the additional facial features of one of the two similar faces, it does not affect the final face recognition effect.

[0087] Therefore, the data storage method of the face database provided in this embodiment can not only save the storage resources of the face database, but also easily distinguish two similar faces, thereby avoiding the technical defects of misidentification in the face recognition process and improving the accuracy of face recognition.

[0088] In one embodiment, such as Figure 6 As shown, step 101 above includes steps 601 to 603, wherein:

[0089] Step 601: Obtain the second similarity between the first basic facial feature corresponding to the face image to be identified and the second basic facial feature corresponding to the face of the currently registered user.

[0090] The second similarity represents the similarity between the first and second basic facial features.

[0091] Step 602: If the second similarity is greater than a preset threshold, the face of the currently registered user is identified as the face of the candidate registered user.

[0092] Step 603: If the second similarity is determined to be no greater than a preset threshold, repeat the above steps based on the face of the next registered user until the face of the candidate registered user is determined.

[0093] Specifically, if the second similarity is determined to be no greater than a preset threshold, the second similarity between the first basic facial features corresponding to the face image to be identified and the second basic facial features corresponding to the face of the next registered user is obtained. If the second similarity is determined to be greater than the preset threshold, the face of the next registered user is identified as a candidate registered user face.

[0094] Steps 601 to 603 above determine whether the facial feature similarity between the current registered user's face and the face image to be identified is greater than a preset threshold. If the current registered user's face matches the face image to be identified, the matching process stops, and the current registered user's face is identified as a candidate registered user's face. If the current registered user's face does not match the face image to be identified, the process continues to determine whether the next registered user's face matches the face image to be identified, and so on, until a candidate registered user's face that matches the face image to be identified is identified. Since the above matching process only matches one registered user's face at a time, and stops as soon as a match is successful, it is not necessary to traverse the entire face database to match a suitable candidate registered user's face, thereby saving the processor's data processing resources and improving the processor's data processing efficiency.

[0095] The face recognition device provided by the present invention will be described below. The face recognition device described below can be referred to in correspondence with the face recognition method described above.

[0096] like Figure 7 As shown, the present invention provides a face recognition device, the face recognition device 100 comprising:

[0097] The face matching module 101 is used to determine the candidate registered user face that matches the face image to be identified based on the first basic face feature corresponding to the face image to be identified and the second basic face feature corresponding to each registered user face in the face database.

[0098] The similarity comparison module 102 is used to obtain the first facial feature similarity between the candidate registered user's face and the face image to be identified, and the second facial feature similarity between the similar registered user's face and the face image to be identified, when it is determined that there is a similar registered user's face among the candidate registered user's faces.

[0099] The face recognition module 103 is used to determine the target registered user face corresponding to the face image to be recognized from the candidate registered user faces and similar registered user faces based on the numerical comparison results of the first face feature similarity and the second face feature similarity.

[0100] In one embodiment, the similarity comparison module 102 includes a first similarity calculation unit, configured to determine a first face feature similarity based on the first additional face feature corresponding to the face image to be identified and the second additional face feature corresponding to the face of the candidate registered user when additional face features corresponding to the face of the candidate registered user exist in the face database; and to determine a first face feature similarity based on the first basic face feature corresponding to the face image to be identified and the second basic face feature corresponding to the face of the candidate registered user when additional face features corresponding to the face of the candidate registered user do not exist in the face database, wherein the feature point density of the additional face features is greater than the feature point density of the basic face features.

[0101] In one embodiment, the similarity comparison module 102 further includes a second similarity calculation unit, used to determine the second face feature similarity based on the first additional face feature corresponding to the face image to be identified and the third additional face feature corresponding to the face of the similar registered user when there are additional face features corresponding to the face of the similar registered user in the face database; and to determine the second face feature similarity based on the first basic face feature corresponding to the face image to be identified and the third basic face feature corresponding to the face of the similar registered user when there are no additional face features corresponding to the face of the similar registered user in the face database, wherein the feature point density of the additional face features is greater than the feature point density of the basic face features.

[0102] In one embodiment, the face recognition module 103 is further configured to: determine the candidate registered user face as the target registered user face corresponding to the face image to be recognized when the first face feature similarity is greater than the second face feature similarity; and determine the similar registered user face as the target registered user face corresponding to the face image to be recognized when the second face feature similarity is greater than the first face feature similarity.

[0103] In one embodiment, the face recognition device 100 further includes a face database module, used to obtain a first similarity between the fourth basic face feature corresponding to the face image to be added to the database and the second basic face feature corresponding to the face of a registered user in the face database; if the first similarity is greater than a preset threshold, to construct a similarity association between the face image to be added to the database and the face of the registered user; to extract the fourth additional face feature corresponding to the face image to be added to the database, and to store the fourth basic face feature and the fourth additional face feature corresponding to the face image to be added to the database into the face database.

[0104] In one embodiment, the face matching module 101 is further configured to obtain a second similarity between a first basic face feature corresponding to the face image to be identified and a second basic face feature corresponding to the face of the currently registered user; if the second similarity is determined to be greater than a preset threshold, the face of the currently registered user is determined as a candidate registered user face; if the second similarity is determined to be less than the preset threshold, the above steps are repeated based on the next registered user face until a candidate registered user face is determined.

[0105] Figure 8 An example is a schematic diagram of the physical structure of an electronic device, such as... Figure 8 As shown, the electronic device may include a processor 810, a communications interface 820, a memory 830, and a communication bus 840, wherein the processor 810, the communications interface 820, and the memory 830 communicate with each other through the communication bus 840. The processor 810 can call logical instructions in the memory 830 to execute the face recognition method provided by the above methods. The method includes: determining candidate registered user faces that match the face image to be recognized based on a first basic face feature corresponding to the face image to be recognized and a second basic face feature corresponding to each registered user face in the face database; if it is determined that there are similar registered user faces among the candidate registered user faces, obtaining the first face feature similarity between the candidate registered user face and the face image to be recognized, and the second face feature similarity between the similar registered user face and the face image to be recognized; and determining the target registered user face corresponding to the face image to be recognized from the candidate registered user faces and the similar registered user faces based on the numerical comparison results of the first face feature similarity and the second face feature similarity.

[0106] Furthermore, the logical instructions in the aforementioned memory 830 can be implemented as software functional units and, when sold or used as independent products, can be stored in a computer-readable storage medium. Based on this understanding, the technical solution of the present invention, in essence, or the part that contributes to the prior art, or a part of the technical solution, can be embodied in the form of a software product. This computer software product is stored in a storage medium and includes several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute all or part of the steps of the methods of the various embodiments of the present invention. The aforementioned storage medium includes various media capable of storing program code, such as USB flash drives, portable hard drives, read-only memory (ROM), random access memory (RAM), magnetic disks, or optical disks.

[0107] On the other hand, the present invention also provides a computer program product, which includes a computer program that can be stored on a computer-readable storage medium. When the computer program is executed by a processor, the computer is able to execute the face recognition method provided by the above methods. The method includes: determining candidate registered user faces that match the face image to be recognized based on a first basic face feature corresponding to the face image to be recognized and a second basic face feature corresponding to each registered user face in the face database; when it is determined that there are similar registered user faces among the candidate registered user faces, obtaining a first face feature similarity between the candidate registered user face and the face image to be recognized, and a second face feature similarity between the similar registered user face and the face image to be recognized; and determining the target registered user face corresponding to the face image to be recognized from the candidate registered user faces and the similar registered user faces based on the numerical comparison result of the first face feature similarity and the second face feature similarity.

[0108] In another aspect, the present invention also provides a computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the face recognition method provided by the above methods. The method includes: determining candidate registered user faces that match the face image to be recognized based on a first basic face feature corresponding to a face image to be recognized and a second basic face feature corresponding to each registered user face in a face database; if it is determined that there are similar registered user faces among the candidate registered user faces, obtaining a first face feature similarity between the candidate registered user face and the face image to be recognized, and a second face feature similarity between the similar registered user face and the face image to be recognized; and determining the target registered user face corresponding to the face image to be recognized from the candidate registered user faces and the similar registered user faces based on the numerical comparison result of the first face feature similarity and the second face feature similarity.

[0109] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units; that is, they may be located in one place or distributed across multiple network units. Some or all of the modules can be selected to achieve the purpose of this embodiment according to actual needs. Those skilled in the art can understand and implement this without any creative effort.

[0110] Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus necessary general-purpose hardware platforms, and of course, it can also be implemented by hardware. Based on this understanding, the above technical solutions, in essence or the part that contributes to the prior art, can be embodied in the form of a software product. This computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, magnetic disk, optical disk, etc., including several instructions to cause a computer device (which may be a personal computer, server, or network device, etc.) to execute the methods of various embodiments or some parts of embodiments.

[0111] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, and not to limit them; although the present invention has been described in detail with reference to the foregoing embodiments, those skilled in the art should understand that modifications can still be made to the technical solutions described in the foregoing embodiments, or equivalent substitutions can be made to some of the technical features; and these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A face recognition method, characterized in that, include: Based on the first basic facial features corresponding to the face image to be identified and the second basic facial features corresponding to the faces of each registered user in the face database, candidate registered user faces that match the face image to be identified are determined. If it is determined that the candidate registered user's face has a similar registered user's face, the first facial feature similarity between the candidate registered user's face and the face image to be identified, and the second facial feature similarity between the similar registered user's face and the face image to be identified are obtained. Based on the numerical comparison results of the first facial feature similarity and the second facial feature similarity, the target registered user face corresponding to the face image to be identified is determined from the candidate registered user face and the similar registered user face; The step of obtaining the first facial feature similarity between the candidate registered user's face and the face image to be identified includes: If the candidate registered user's face has additional facial features in the face database, the similarity of the first facial features is determined based on the first additional facial features corresponding to the face image to be identified and the second additional facial features corresponding to the candidate registered user's face. If no additional facial features corresponding to the candidate registered user's face exist in the face database, the similarity of the first facial features is determined based on the first basic facial features corresponding to the face image to be identified and the second basic facial features corresponding to the candidate registered user's face. The feature point density of the additional facial features is greater than the feature point density of the basic facial features. The step of obtaining the similarity of the second facial features between the similar registered user's face and the face image to be identified includes: If the face database contains additional face features corresponding to the similar registered user's face, the second face feature similarity is determined based on the first additional face feature corresponding to the face image to be identified and the third additional face feature corresponding to the similar registered user's face. If no additional facial features corresponding to the similar registered user's face exist in the face database, the second facial feature similarity is determined based on the first basic facial features corresponding to the face image to be identified and the third basic facial features corresponding to the similar registered user's face. The feature point density of the additional facial features is greater than the feature point density of the basic facial features.

2. The face recognition method according to claim 1, characterized in that, The step of determining the target registered user face corresponding to the face image to be identified from the candidate registered user faces and the similar registered user faces based on the numerical comparison results of the first facial feature similarity and the second facial feature similarity includes: If the similarity of the first facial feature is greater than that of the second facial feature, the candidate registered user face is determined as the target registered user face corresponding to the face image to be identified; If the similarity of the second facial feature is greater than that of the first facial feature, the similar registered user face is identified as the target registered user face corresponding to the face image to be identified.

3. The face recognition method according to claim 1, characterized in that, The method further includes: Obtain the first similarity between the fourth basic facial feature corresponding to the face image to be added to the database and the second basic facial feature corresponding to the face of a registered user in the face database; If the first similarity is greater than a preset threshold, a similarity association relationship is constructed between the face image to be added to the database and the face of the registered user. Extract the fourth additional facial feature corresponding to the face image to be added to the database, and store the fourth basic facial feature and the fourth additional facial feature corresponding to the face image to be added to the database.

4. The face recognition method according to claim 1, characterized in that, The step of determining candidate registered user faces that match the face image to be identified, based on the first basic facial features corresponding to the face image to be identified and the second basic facial features corresponding to the faces of each registered user in the face database, includes: Obtain the second similarity between the first basic facial features corresponding to the face image to be identified and the second basic facial features corresponding to the face of the currently registered user; If the second similarity is determined to be greater than a preset threshold, the face of the currently registered user is determined as the face of the candidate registered user; If the second similarity is determined to be no greater than a preset threshold, the above steps are repeated based on the face of the next registered user until the face of the candidate registered user is determined.

5. A face recognition device, characterized in that, include: The face matching module is used to determine candidate registered user faces that match the face image to be identified based on the first basic face features corresponding to the face image to be identified and the second basic face features corresponding to the faces of each registered user in the face database. The similarity comparison module is used to obtain, when it is determined that there is a similar registered user face to the candidate registered user face, the first facial feature similarity between the candidate registered user face and the face image to be identified, and the second facial feature similarity between the similar registered user face and the face image to be identified; A face recognition module is used to determine the target registered user face corresponding to the face image to be recognized from the candidate registered user faces and the similar registered user faces based on the numerical comparison results of the first face feature similarity and the second face feature similarity. The similarity comparison module includes a first similarity calculation unit, which is used to determine the first face feature similarity based on the first additional face feature corresponding to the face image to be identified and the second additional face feature corresponding to the face of the candidate registered user when the additional face feature corresponding to the face of the candidate registered user exists in the face database. If no additional facial features corresponding to the candidate registered user's face exist in the face database, the similarity of the first facial features is determined based on the first basic facial features corresponding to the face image to be identified and the second basic facial features corresponding to the candidate registered user's face. The feature point density of the additional facial features is greater than the feature point density of the basic facial features. The similarity comparison module includes a second similarity calculation unit, which is used to determine the second face feature similarity based on the first additional face feature corresponding to the face image to be identified and the third additional face feature corresponding to the face of the similar registered user when the additional face feature corresponding to the face of the similar registered user exists in the face database. If no additional facial features corresponding to the similar registered user's face exist in the face database, the second facial feature similarity is determined based on the first basic facial features corresponding to the face image to be identified and the third basic facial features corresponding to the similar registered user's face. The feature point density of the additional facial features is greater than the feature point density of the basic facial features.

6. An electronic device comprising a memory, a processor, and a computer program stored in the memory and executable on the processor, characterized in that, When the processor executes the program, it implements the face recognition method as described in any one of claims 1 to 4.

7. A computer-readable storage medium having a computer program stored thereon, characterized in that, When the computer program is executed by the processor, it implements the face recognition method as described in any one of claims 1 to 4.

8. A computer program product, comprising a computer program, characterized in that, When the computer program is executed by the processor, it implements the face recognition method as described in any one of claims 1 to 4.

Images