Face recognition method, device, storage medium and electronic device

Abstract

The present invention provides a face recognition method, device, storage medium and electronic device, comprising obtaining a face video to be recognized, extracting face features and micro-expressions on the face video, and obtaining a first posture feature when the face video has micro-expressions; bringing the face features into a preset three-dimensional face model to construct a three-dimensional face model to be recognized; bringing the micro-expressions into the three-dimensional face model to obtain a second posture feature of a normal face producing micro-expressions; substituting the first posture feature and the second posture feature into a preset big data face analysis model to determine face change information; bringing the face three-dimensional model into change according to the face change information to generate a pre-comparison face set; bringing the face pre-comparison set into a big data face recognition database for matching to determine the face information on the face video.

Description

Technical Field

[0001] The present invention relates to the field of face recognition technology, and in particular to a face recognition method, device, storage medium and electronic equipment. Background Art

[0002] At present, in the field of face recognition, ordinary face recognition has been applied on various electronic devices, but the existing face recognition can only be rough recognition. If the face changes greatly, it cannot be recognized. For example, after committing a crime, a criminal undergoes plastic surgery, and the face after plastic surgery cannot be recognized. Therefore, how to recognize faces with relatively large changes is the research and development direction of technical personnel in this field. Summary of the invention

[0003] The present invention provides a face recognition method, device, storage medium and electronic device, which are used to solve the problem of how to perform face recognition when the face changes greatly, such as plastic surgery.

[0004] A face recognition method, comprising:

[0005] Acquire a face video to be recognized, extract facial features and micro-expressions in the face video, and acquire a first posture feature when the face video has micro-expressions;

[0006] Bringing the facial features into a preset three-dimensional face model to construct a three-dimensional face model to be identified;

[0007] Bringing the micro-expression into the three-dimensional model of the face to be recognized to obtain the second posture feature of the normal face producing the micro-expression;

[0008] Substituting the first posture feature and the second posture feature into a preset big data face analysis model to determine face change information;

[0009] Bringing the face change information into the face three-dimensional model to make changes, and generating a pre-comparison face set;

[0010] The pre-compared face set is brought into a big data face recognition database for matching to determine the face information on the face video.

[0011] As an embodiment of the present invention, the step of obtaining a face video to be identified, extracting face features and micro-expressions in the face video, and obtaining a first posture feature when the face video has micro-expressions includes:

[0012] Divide the face video into frames to obtain a frame image set;

[0013] Extracting facial features of each frame image in the frame image set respectively, wherein the facial features include: facial features, eye features, mouth features, ear features and nose features;

[0014] Extracting frame features of each pixel in the frame image according to the facial features;

[0015] Sorting the frame images according to time to determine the change information of the frame features at different times;

[0016] Substitute the change information of the frame features at different times into the micro-expression database to determine the first posture feature with micro-expression.

[0017] As an embodiment of the present invention, the step of bringing the facial features into a preset three-dimensional face model to construct a three-dimensional face model to be identified includes:

[0018] Acquire the facial features, and determine the three-dimensional attributes of the facial features in the facial video;

[0019] The three-dimensional attributes include position attributes, area attributes and contour attributes;

[0020] Determine, according to the three-dimensional attributes, a required area of ​​the facial feature on the three-dimensional face model;

[0021] Substituting the facial features into the three-dimensional face model according to the area of ​​the required region, and supplementing the facial texture of the non-required region in the three-dimensional face model;

[0022] After the facial texture of the non-required area is supplemented, the three-dimensional model of the face to be recognized is generated.

[0023] As an embodiment of the present invention, the step of bringing the micro-expression into the three-dimensional model of the face to be identified to obtain the second posture feature of the micro-expression produced by the normal face includes:

[0024] Acquire the micro-expression, and determine the expression features of the micro-expression;

[0025] Substituting the expression features into the three-dimensional model of the face to be identified, and simulating the micro-expression through the three-dimensional model of the face to be identified;

[0026] The second posture feature of the human face is determined according to the facial texture movement and muscle movement of the human face when the three-dimensional model of the human face to be recognized simulates the micro-expression.

[0027] As an embodiment of the present invention, the step of substituting the first posture feature and the second posture feature into a preset big data face analysis model to determine face change information includes:

[0028] A preset face database; wherein,

[0029] The face database includes face cosmetic surgery data, face aging data, face disease data and normal face data;

[0030] Preset a big data face analysis model, and connect the face database through the big data face analysis model;

[0031] Substituting the first posture feature and the second posture feature into the big data face analysis model, and determining the distinguishing features of the first posture feature and the second posture feature through comparison processing with the face database;

[0032] According to the distinguishing features, the face change information is determined; wherein,

[0033] The face change information includes muscle change information and face texture change information.

[0034] As an embodiment of the present invention, the step of bringing the face change information into the face 3D model to change it and generating a pre-comparison face set includes:

[0035] Acquire the face change information, and determine the cause of the change corresponding to the face change information;

[0036] The cause of the change is substituted into the three-dimensional face model as a change value, the changed three-dimensional face model is determined, and the face in the face video is restored according to the changed three-dimensional face model to generate a pre-comparison face set.

[0037] As an embodiment of the present invention, the step of bringing the pre-compared face set into a big data face recognition database for matching and determining the face information on the face video includes:

[0038] Step 1: Construct a pre-comparison face matrix based on the pre-comparison face set:

[0039]

[0040] in, represents the pre-comparison face matrix; Indicates The first pre-matched face Facial features of feature points; ; ; Indicates the number of face images in the pre-comparison face combination; The number of facial features representing feature points;

[0041] Step 2: Determine the recognition coefficient of the face recognition model based on the big data face recognition database:

[0042]

[0043] in, It represents the recognition coefficient; Represents the first Personal facial features; Represents the first The corresponding face of each individual facial feature; ; Represents the number of facial features in the big data face recognition database; Represents the recognition credibility value of the big data face recognition database; Represents the recognition weight of the big data face recognition database;

[0044] Step 3: According to the recognition coefficient, match the data of the face matrix to determine the matching value:

[0045]

[0046] in, represents the feature mean;

[0047] Among them, in the , determining the corresponding face features in the big data face recognition database according to the recognition coefficient, and determining the corresponding face image and face identity according to the face features in the big data face recognition database; in the Repeat steps 1 to 3.

[0048] A face recognition device, applied to an electronic device, comprising:

[0049] The first acquisition module is used to acquire a face video to be identified, extract facial features and micro-expressions in the face video, and acquire a first posture feature when the face video has micro-expressions;

[0050] Construction module: used to bring the facial features into a preset three-dimensional face model to construct a three-dimensional face model to be identified;

[0051] The second acquisition module is used to bring the micro-expression into the three-dimensional model of the face to be recognized, and obtain the second posture feature of the normal face producing the micro-expression;

[0052] A first determination module: used for substituting the first posture feature and the second posture feature into a preset big data face analysis model to determine face change information;

[0053] A generation module: used to bring the face change information into the face 3D model for change, and generate a pre-comparison face set;

[0054] The second determination module is used to bring the pre-compared face set into the big data face recognition database for matching, and determine the face information on the face video.

[0055] A storage medium includes a computer program. When the computer program is run on a computer, the computer executes the above-mentioned face recognition method.

[0056] An electronic device comprises a processor and a memory electrically connected to each other, wherein the memory stores a computer program, and the processor is used to execute the above-mentioned face recognition method by calling the computer program.

[0057] The beneficial effect is that the present invention, as a face recognition method, can perform face recognition on people who have undergone plastic surgery, face aging or have suffered serious facial diseases, and its role can be extended to identifying criminals. Compared with the face recognition technology in the prior art, the present invention is a traceable recognition technology. For example, for the recognition of face aging, even if the big data face recognition database contains photos from decades ago, the present invention can also perform step-by-step analysis and processing through face videos to determine the corresponding information of the specific person in the big data face recognition database.

[0058] Other features and advantages of the present invention will be described in the following description, and partly become apparent from the description, or understood by practicing the present invention. The purpose and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description and the accompanying drawings.

[0059] The technical solution of the present invention is further described in detail below through the accompanying drawings and embodiments. BRIEF DESCRIPTION OF THE DRAWINGS

[0060] The accompanying drawings are used to provide a further understanding of the present invention and constitute a part of the specification. Together with the embodiments of the present invention, they are used to explain the present invention and do not constitute a limitation of the present invention. In the accompanying drawings:

[0061] Figure 1 A method flow chart of a face recognition method in an embodiment of the present invention;

[0062] Figure 2 2 is a diagram showing the composition of a face recognition device in an embodiment of the present invention. DETAILED DESCRIPTION

[0063] The preferred embodiments of the present invention are described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention, and are not used to limit the present invention.

[0064] In the following description, the specific embodiments of the present application will be described with reference to the steps and symbols performed by one or more computers, unless otherwise stated. Therefore, these steps and operations will be mentioned several times as being performed by a computer, and the computer execution referred to herein includes the operation of a computer processing unit by an electronic signal representing data in a structured form. This operation converts the data or maintains it at a location in the memory system of the computer, which can be reconfigured or otherwise change the operation of the computer in a manner familiar to testers in the field. The data structure maintained by the data is a physical location in the memory, which has specific characteristics defined by the data format. However, the principles of the present application are described in the above text, which does not represent a limitation, and testers in the field will understand that the various steps and operations described below can also be implemented in hardware.

[0065] The term "module" used in this article can be regarded as a software object executed on the computing system. The different components, modules, engines and services described in this article can be regarded as implementation objects on the computing system. The device and method described in this article can be implemented in software, and of course can also be implemented on hardware, all within the scope of protection of this application. The terms "first", "second" and "third" in this application are used to distinguish different objects, rather than to describe a specific order. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions. For example, a process, method, system, product or device that includes a series of steps or modules is not limited to the listed steps or modules, but some embodiments also include steps or modules that are not listed, or some embodiments also include other steps or modules inherent to these processes, methods, products or devices.

[0066] As attached Figure 1 As shown, the present invention is a face recognition method, comprising:

[0067] Acquire a face video to be recognized, extract facial features and micro-expressions in the face video, and acquire a first posture feature when the face video has micro-expressions;

[0068] The face video to be identified in the present invention means that when performing face recognition again, the present invention is to record the face of the user to be identified within a period of time or a preset time. The present invention cannot directly use pictures because the micro-expression features of the user in the picture cannot be fully discovered. The first posture feature represents the features of the entire face of the user with micro-expressions, including facial features, eye features, mouth features, ear features and nose features.

[0069] The facial features are brought into a preset three-dimensional face model to construct a three-dimensional model of the face to be identified; the preset three-dimensional face model is a simulation architecture model of a general human head made by technical personnel in this field through three-dimensional simulation of three-dimensional simulation software, but after the facial features are substituted, its display state is the same as the face in the face video.

[0070] The micro-expression is brought into the three-dimensional model of the face to be identified, and the second posture feature of the normal face producing micro-expression is obtained; the present invention substitutes the micro-expression into the three-dimensional model of the face to be identified because the simulation structure model made of the general human head can only be suitable for normal people. For people who have undergone plastic surgery, there is an incoordination on their faces. Therefore, their micro-expression state is different from that of normal faces, especially the texture of muscles and faces is different. Therefore, although the micro-expression displayed by the second posture feature is the same as the micro-expression on the face video, when implementing the micro-expression, its muscles and facial texture may not meet the level that normal people can achieve.

[0071] Substituting the first posture feature and the second posture feature into a preset big data face analysis model to determine face change information;

[0072] The big data face analysis model of the present invention is connected to a face database, which includes face cosmetic surgery data, face aging data, face disease data and normal face data; therefore, through the difference in posture features, the texture of the muscles and the face in the second posture feature does not conform to that of a normal person, but conforms to any one of the cosmetic surgery data, face aging data and face disease data, it can be judged that the micro-expression of the face is in a state that a normal face cannot make due to any one of factors such as cosmetic surgery data, face aging, face disease, etc.

[0073] The face change information is brought into the face three-dimensional model for change to generate a pre-comparison face set; the face change information is any one of plastic surgery, face aging or face disease. After it is substituted into the face three-dimensional model, the appearance of the human face before these factors are adjusted out can be determined, but because there may be multiple face change information that will lead to the same result (a certain micro-expression can be achieved through multiple body change information), the present invention obtains a pre-recognized face set with multiple possibilities.

[0074] The pre-compared face set is brought into the big data face recognition database for matching, and the face information on the face video is determined. Because it is impossible for two people to be exactly the same, even twins have different micro-expressions, so the present invention uses data portrait matching in the last step. The faces in the pre-identified face set are matched with the face images in the big data face recognition database to confirm the final face. The face data in the big data face recognition database is obtained through the household registration system, the public security system, and other institutions that have the right to collect citizen information. When there are no relevant persons who can be matched, it can be treated as unrecognizable. The unrecognizable person must be a criminal who has not been registered. As a face recognition method, the present invention can perform face recognition on people who have undergone plastic surgery, face aging or have suffered serious facial diseases. Its role can be extended to identifying criminals. Compared with the face recognition technology in the prior art, the present invention is a recognition technology that can be traced back. For example, for the recognition of face aging, even if the big data face recognition database contains photos from decades ago, the present invention can also perform step-by-step analysis and processing through face videos to determine the corresponding information of the specific person in the big data face recognition database.

[0075] As an embodiment of the present invention, the step of obtaining a face video to be identified, extracting face features and micro-expressions in the face video, and obtaining a first posture feature when the face video has micro-expressions includes:

[0076] Divide the face video into frames to obtain a frame image set;

[0077] The present invention divides the face video into frames, and obtains a frame image set after the frame division. The frame image set is divided based on time, and each moment has a frame of image.

[0078] The facial features of each frame image in the frame image set are extracted respectively, wherein the facial features include: facial features, eye features, mouth features, ear features and nose features; in the present invention, the definition of the features of the facial features, eye features, mouth features, ear features and nose features is extended to the pixel level, and the features of all pixels are comprehensively divided into facial features, eye features, mouth features, ear features and nose features.

[0079] Extracting frame features of each pixel in the frame image according to the facial features;

[0080] The frame images are sorted according to time to determine the change information of the frame features at different times; the change information of different frames reflects the changes in the user's expression, especially the changes in micro-expressions. Through cosmetic surgery technology, the human face can change, but the long-term habits of micro-expressions cannot change.

[0081] Substitute the change information of the frame features at different times into the micro-expression database to determine the first posture feature with micro-expression. The present invention can realize accurate recognition of human face by extracting micro-expressions and facial features, and more data is obtained based on the extraction of micro-expressions and facial features from frame images, which can be more accurate in face recognition.

[0082] As an embodiment of the present invention, the step of bringing the facial features into a preset three-dimensional face model to construct a three-dimensional face model to be identified includes:

[0083] Acquire the facial features, and determine the three-dimensional attributes of the facial features in the facial video;

[0084] The three-dimensional attributes include position attributes (mainly spatial position), area attributes (attributes of multiple faces) and contour attributes (the overall contour includes length, width and height); when extracting attributes, the present invention will establish a three-dimensional space coordinate system, and based on the three-dimensional space coordinate system, accurately extract the three-dimensional attributes.

[0085] According to the three-dimensional attributes, the required area of ​​the facial features on the three-dimensional face model is determined; that is, the area occupied by the common facial features in the face model is the required area. The required area is the area occupied by the face, facial features, chin and other parts of the human body.

[0086] According to the area of ​​the required region, the facial features are substituted into the three-dimensional face model, and the facial texture of the non-required region in the three-dimensional face model is supplemented; the non-required region refers to the neck connected to the chin, the top of the head and other parts.

[0087] After the facial texture of the non-required area is supplemented, the three-dimensional model of the face to be identified is generated. The generated face model to be identified is an overall face model.

[0088] As an embodiment of the present invention, the step of bringing the micro-expression into the three-dimensional model of the face to be identified to obtain the second posture feature of the micro-expression produced by the normal face includes:

[0089] The micro-expression is obtained, and the expression feature of the micro-expression is determined; the expression feature of the micro-expression is the texture change feature of the face when the human body produces the micro-expression.

[0090] Substituting the expression features into the three-dimensional model of the face to be identified, and simulating the micro-expression through the three-dimensional model of the face to be identified;

[0091] The second posture feature of the human face is determined according to the facial texture movement and muscle movement of the human face when the three-dimensional model of the human face to be recognized simulates the micro-expression. The facial texture movement is the embodiment of the micro-expression, and the muscle movement is the power source of the micro-expression change.

[0092] As an embodiment of the present invention, the step of substituting the first posture feature and the second posture feature into a preset big data face analysis model to determine face change information includes:

[0093] A preset face database; wherein,

[0094] The face database includes face cosmetic surgery data, face aging data, face disease data and normal face data;

[0095] A big data face analysis model is preset, and the face database is connected through the big data face analysis model; the big data face analysis model is obtained by improving parameters of a general traceability analysis model.

[0096] The first posture feature and the second posture feature are substituted into the big data face analysis model, and the distinguishing features of the first posture feature and the second posture feature are determined through comparison processing of the face database; the comparison processing can obtain the differences in posture features, and these differences are only caused by plastic surgery, facial diseases, aging or other reasons, such as facial deformity, while a normal face will not produce such postures.

[0097] According to the distinguishing features, the face change information is determined; wherein,

[0098] The face change information includes muscle change information and face texture change information.

[0099] As an embodiment of the present invention, the step of bringing the face change information into the face 3D model to change it and generating a pre-comparison face set includes:

[0100] The face change information is obtained, and the cause of the change corresponding to the face change information is determined; the cause corresponding to the face change information is factors such as plastic surgery, face aging or disease.

[0101] Substitute the change cause as the change value into the face 3D model, determine the changed face 3D model, and restore the face in the face video according to the changed face 3D model to generate a pre-comparison face set. After the change value of the change cause is imported into the face 3D model, the face in the face video will be restored according to the change cause. For example, if the face is aged, it will be restored to a younger state, and if the face has undergone plastic surgery, it will be restored to the state before the plastic surgery. Because the cause is known, it is easier to restore. Because the cause of the change may be multi-factorial, a pre-comparison face set is generated.

[0102] As an embodiment of the present invention, the step of bringing the pre-compared face set into a big data face recognition database for matching and determining the face information on the face video includes:

[0103] Step 1: Construct a pre-comparison face matrix based on the pre-comparison face set:

[0104]

[0105] in, represents the pre-comparison face matrix; Indicates The first pre-matched face Facial features of feature points; ; ; Indicates the number of face images in the pre-comparison face combination; The number of facial features representing feature points;

[0106] Step 2: Determine the recognition coefficient of the face recognition model based on the big data face recognition database:

[0107]

[0108] in, It represents the recognition coefficient; Represents the first Personal facial features; Represents the first The corresponding face of each individual facial feature; ; Represents the number of facial features in the big data face recognition database; Represents the recognition credibility value of the big data face recognition database; Represents the recognition weight of the big data face recognition database;

[0109] Step 3: According to the recognition coefficient, match the data of the face matrix to determine the matching value:

[0110]

[0111] in, represents the feature mean;

[0112] Among them, in the , determining the corresponding face features in the big data face recognition database according to the recognition coefficient, and determining the corresponding face image and face identity according to the face features in the big data face recognition database; in the Repeat steps 1 to 3.

[0113] The principle and beneficial effect of the above technical solution are as follows: in order to confirm the facial information in the final confirmation link by data matching, compared with the matching method of the prior art, the present invention pre-constructs a pre-compared face matrix, in which all the features of the pre-identified face have been statistically analyzed, thereby ensuring the correctness and integrity of the recognition data. Then, the calculation of the recognition coefficient is realized based on the big data face recognition database. The recognition coefficient of the big data face recognition database combines all the facial features in the big data face recognition database and is based on the recognized credible value, which is obtained through machine learning based on a large amount of data. Finally, the present invention determines the final recognized face based on the matching relationship between the recognition coefficient and the face matrix in the face matrix. And retrieve the identity information corresponding to the recognized face.

[0114] As attached Figure 2 As shown, a face recognition device is applied to an electronic device, comprising:

[0115] The first acquisition module is used to acquire a face video to be identified, extract facial features and micro-expressions in the face video, and acquire a first posture feature when the face video has micro-expressions;

[0116] Construction module: used to bring the facial features into a preset three-dimensional face model to construct a three-dimensional face model to be identified;

[0117] The second acquisition module is used to bring the micro-expression into the three-dimensional model of the face to be recognized, and obtain the second posture feature of the normal face producing the micro-expression;

[0118] A first determination module: used for substituting the first posture feature and the second posture feature into a preset big data face analysis model to determine face change information;

[0119] A generation module: used to bring the face change information into the face 3D model for change, and generate a pre-comparison face set;

[0120] The second determination module is used to bring the pre-compared face set into the big data face recognition database for matching, and determine the face information on the face video. As a face recognition method, the present invention can perform face recognition on people who have undergone plastic surgery, face aging, or have suffered major facial diseases. Its role can be extended to the identification of criminals. Compared with the face recognition technology in the prior art, the present invention belongs to a traceable identification technology. For example, for the identification of aging faces, even if the big data face recognition database contains photos from decades ago, the present invention can also perform step-by-step analysis and processing through face videos to determine the corresponding information of the specific person in the big data face recognition database.

[0121] A storage medium, the storage medium including a computer program, when the computer program is run on a computer, enables the computer to perform the above-mentioned face recognition method. In some embodiments, the storage medium can be a disk, an optical disk, a read-only memory (ROM), or a random access memory (RAM).

[0122] An electronic device comprises a processor and a memory electrically connected to each other, wherein the memory stores a computer program, and the processor is used to execute the above-mentioned face recognition method by calling the computer program.

[0123] Obviously, those skilled in the art can make various changes and modifications to the present invention without departing from the spirit and scope of the present invention. Thus, if these modifications and variations of the present invention fall within the scope of the claims of the present invention and their equivalents, the present invention is also intended to include these modifications and variations.

Claims

1. A face recognition method, characterized in that: include: Acquire a face video to be recognized, extract facial features and micro-expressions in the face video, and acquire a first posture feature when the face video has micro-expressions; Bringing the facial features into a preset three-dimensional face model to construct a three-dimensional face model to be identified; Bringing the micro-expression into the three-dimensional model of the face to be recognized to obtain the second posture feature of the normal face producing the micro-expression; Substituting the first posture feature and the second posture feature into a preset big data face analysis model to determine face change information; Bringing the face change information into the face three-dimensional model to make changes, and generating a pre-comparison face set; Bringing the pre-compared face set into a big data face recognition database for matching, and determining the face information on the face video; The pre-comparison face set is brought into a big data face recognition database for matching, and the face information on the face video is determined to include: Step 1: constructing a pre-comparison face matrix according to the pre-comparison face set; Step 2: Determine the recognition coefficient of the face recognition model based on the big data face recognition database; Step 3: According to the recognition coefficient, match the data of the face matrix to determine the matching value.

2. A face recognition method as claimed in claim 1, characterized in that: The step of obtaining a face video to be identified, extracting face features and micro-expressions in the face video, and obtaining a first posture feature when the face video has micro-expressions includes: Divide the face video into frames to obtain a frame image set; Extracting facial features of each frame image in the frame image set respectively, wherein the facial features include: facial features, eye features, mouth features, ear features and nose features; Extracting frame features of each pixel in the frame image according to the facial features; Sorting the frame images according to time to determine the change information of the frame features at different times; Substitute the change information of the frame features at different times into the micro-expression database to determine the first posture feature with micro-expression.

3. A face recognition method as claimed in claim 1, characterized in that: The step of bringing the facial features into a preset three-dimensional face model to construct a three-dimensional face model to be identified includes: Acquire the facial features, and determine the three-dimensional attributes of the facial features in the facial video; The three-dimensional attributes include position attributes, area attributes and contour attributes; Determine, according to the three-dimensional attributes, a required area of ​​the facial feature on the three-dimensional face model; Substituting the facial features into the three-dimensional face model according to the area of ​​the required region, and supplementing the facial texture of the non-required region in the three-dimensional face model; After the facial texture of the non-required area is supplemented, the three-dimensional model of the face to be recognized is generated.

4. A face recognition method as claimed in claim 1, characterized in that: The step of bringing the micro-expression into the three-dimensional model of the face to be identified to obtain a second posture feature of a normal face producing micro-expression includes: Acquire the micro-expression, and determine the expression features of the micro-expression; Substituting the expression features into the three-dimensional model of the face to be identified, and simulating the micro-expression through the three-dimensional model of the face to be identified; The second posture feature of the human face is determined according to the facial texture movement and muscle movement of the human face when the three-dimensional model of the human face to be recognized simulates micro-expressions.

5. A face recognition method as claimed in claim 1, characterized in that: Substituting the first posture feature and the second posture feature into a preset big data face analysis model to determine face change information includes: A preset face database; wherein, The face database includes face cosmetic surgery data, face aging data, face disease data and normal face data; Preset a big data face analysis model, and connect the face database through the big data face analysis model; Substituting the first posture feature and the second posture feature into the big data face analysis model, and determining the distinguishing features of the first posture feature and the second posture feature through comparison processing with the face database; According to the distinguishing features, the face change information is determined; wherein, The face change information includes muscle change information and face texture change information.

6. A face recognition method as claimed in claim 1, characterized in that: Bringing the face change information into the face 3D model to change it and generate a pre-comparison face set, including: Acquire the face change information, and determine the cause of the change corresponding to the face change information; The cause of the change is substituted into the three-dimensional face model as a change value, the changed three-dimensional face model is determined, and the face in the face video is restored according to the changed three-dimensional face model to generate a pre-comparison face set.

7. A face recognition method as claimed in claim 1, characterized in that: Bringing the pre-compared face set into a big data face recognition database for matching, and determining the face information on the face video, including: Step 1: Construct a pre-comparison face matrix based on the pre-comparison face set: in, represents the pre-comparison face matrix; Indicates The first pre-matched face Facial features of feature points; ; ; represents the number of face images in the pre-comparison face combination; The number of facial features representing feature points; Step 2: Determine the recognition coefficient of the face recognition model based on the big data face recognition database: in, It represents the recognition coefficient; Represents the first Personal facial features; Represents the first The corresponding human face of each facial feature; ; Represents the number of facial features in the big data face recognition database; Represents the recognition credibility value of the big data face recognition database; Represents the recognition weight of the big data face recognition database; Step 3: According to the recognition coefficient, match the data of the face matrix to determine the matching value: in, represents the feature mean; Among them, in the , determining the corresponding face features in the big data face recognition database according to the recognition coefficient, and determining the corresponding face image and face identity according to the face features in the big data face recognition database; in the Repeat steps 1 to 3.

8. A face recognition device, applied to an electronic device, for executing any one of the methods 1 to 7, characterized in that: include: The first acquisition module is used to acquire a face video to be identified, extract facial features and micro-expressions in the face video, and acquire a first posture feature when the face video has micro-expressions; Construction module: used to bring the facial features into a preset three-dimensional face model to construct a three-dimensional face model to be identified; The second acquisition module is used to bring the micro-expression into the three-dimensional model of the face to be recognized, and obtain the second posture feature of the normal face producing the micro-expression; A first determination module: used for substituting the first posture feature and the second posture feature into a preset big data face analysis model to determine face change information; A generation module: used to bring the face change information into the face 3D model for change, and generate a pre-comparison face set; The second determination module is used to bring the pre-compared face set into the big data face recognition database for matching, and determine the face information on the face video.

9. A storage medium, characterized in that: The storage medium includes a computer program, and when the computer program is executed on a computer, the computer is enabled to execute the face recognition method according to any one of claims 1 to 7.

10. An electronic device, characterized in that: It comprises a processor and a memory electrically connected to each other, the memory stores a computer program, and the processor is used to execute the face recognition method according to any one of claims 1 to 7 by calling the computer program.

Images