CN114511911A - Face recognition method, device and equipment - Google Patents
Face recognition method, device and equipment Download PDFInfo
- Publication number
- CN114511911A CN114511911A CN202210181933.XA CN202210181933A CN114511911A CN 114511911 A CN114511911 A CN 114511911A CN 202210181933 A CN202210181933 A CN 202210181933A CN 114511911 A CN114511911 A CN 114511911A
- Authority
- CN
- China
- Prior art keywords
- face
- data
- user
- face data
- parameterized
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000000034 method Methods 0.000 title claims abstract description 66
- 238000006243 chemical reaction Methods 0.000 claims abstract description 23
- 230000000875 corresponding effect Effects 0.000 claims description 72
- 238000001514 detection method Methods 0.000 claims description 26
- 230000008859 change Effects 0.000 claims description 17
- 238000012545 processing Methods 0.000 claims description 17
- 230000009467 reduction Effects 0.000 claims description 10
- 239000013598 vector Substances 0.000 claims description 5
- 230000002596 correlated effect Effects 0.000 claims description 3
- 239000000126 substance Substances 0.000 claims description 3
- 238000010586 diagram Methods 0.000 description 28
- 230000008569 process Effects 0.000 description 22
- 230000006870 function Effects 0.000 description 10
- 230000006872 improvement Effects 0.000 description 9
- 238000004590 computer program Methods 0.000 description 7
- 230000000694 effects Effects 0.000 description 7
- 238000003384 imaging method Methods 0.000 description 7
- 238000005516 engineering process Methods 0.000 description 6
- 238000004364 calculation method Methods 0.000 description 5
- 238000012216 screening Methods 0.000 description 4
- 238000012549 training Methods 0.000 description 4
- 230000002146 bilateral effect Effects 0.000 description 3
- 230000001680 brushing effect Effects 0.000 description 3
- 210000000887 face Anatomy 0.000 description 3
- 238000001914 filtration Methods 0.000 description 3
- 238000012935 Averaging Methods 0.000 description 2
- 238000009825 accumulation Methods 0.000 description 2
- 230000008901 benefit Effects 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000013527 convolutional neural network Methods 0.000 description 2
- 238000013135 deep learning Methods 0.000 description 2
- 230000007613 environmental effect Effects 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000003062 neural network model Methods 0.000 description 2
- 230000003287 optical effect Effects 0.000 description 2
- 238000007781 pre-processing Methods 0.000 description 2
- 238000009877 rendering Methods 0.000 description 2
- 230000009466 transformation Effects 0.000 description 2
- 208000027418 Wounds and injury Diseases 0.000 description 1
- 230000009471 action Effects 0.000 description 1
- 230000003044 adaptive effect Effects 0.000 description 1
- 230000004075 alteration Effects 0.000 description 1
- 238000004458 analytical method Methods 0.000 description 1
- 230000009286 beneficial effect Effects 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 230000001413 cellular effect Effects 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 230000006378 damage Effects 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 210000001061 forehead Anatomy 0.000 description 1
- 238000009499 grossing Methods 0.000 description 1
- 208000014674 injury Diseases 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000011159 matrix material Substances 0.000 description 1
- 230000005055 memory storage Effects 0.000 description 1
- 229920001296 polysiloxane Polymers 0.000 description 1
- 230000000750 progressive effect Effects 0.000 description 1
- 210000001747 pupil Anatomy 0.000 description 1
- 230000003068 static effect Effects 0.000 description 1
- 238000012795 verification Methods 0.000 description 1
Images
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/21—Design or setup of recognition systems or techniques; Extraction of features in feature space; Blind source separation
- G06F18/214—Generating training patterns; Bootstrap methods, e.g. bagging or boosting
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
Landscapes
- Engineering & Computer Science (AREA)
- Data Mining & Analysis (AREA)
- Theoretical Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Bioinformatics & Cheminformatics (AREA)
- Bioinformatics & Computational Biology (AREA)
- Artificial Intelligence (AREA)
- Evolutionary Biology (AREA)
- Evolutionary Computation (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Life Sciences & Earth Sciences (AREA)
- Image Analysis (AREA)
Abstract
The embodiment of the specification discloses a face recognition method, a face recognition device and face recognition equipment. The scheme comprises the following steps: after the face of the user is at a preset face placing position, acquiring a 3D face image based on depth information aiming at the face of the user; performing coordinate conversion on the 3D face image to obtain a 3D face point cloud corresponding to the face of the user; predicting the fitting parameters required by 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud; according to the fitting parameters, fitting a preset 3D reference face grid to obtain 3D parametric face data, wherein the 3D reference face grid comprises a plurality of vertexes, the vertexes are connected, and face semantics are respectively arranged on the vertexes; and identifying the user according to the 3D parameterized face data.
Description
Technical Field
The present disclosure relates to the field of computer technologies, and in particular, to a method, an apparatus, and a device for face recognition.
Background
With the development of computer and internet technologies, many businesses use face recognition to perform identity verification on users, such as face-brushing payment and face-brushing access control.
Face recognition typically includes two forms, 2D face recognition and 3D face recognition. For 3D face recognition, a 3D camera is required to collect a 3D face image of a user, and the face recognition is performed on the user through the 3D face image. The 3D camera includes various types, such as a structured light 3D camera, a ToF3D camera, and the like. Taking a structured light 3D camera which is widely applied as an example, in some special scenes (for example, a strong light environment and when a user exceeds a camera range), the imaging effect of a 3D face image acquired by the structured light 3D camera is often poor.
Based on this, a more reliable face recognition scheme is needed.
Disclosure of Invention
One or more embodiments of the present disclosure provide a face recognition method, apparatus, device, and storage medium, so as to solve the following technical problems: a more reliable face recognition scheme is needed.
To solve the above technical problem, one or more embodiments of the present specification are implemented as follows:
one or more embodiments of the present specification provide a face recognition method, including:
after the face of a user is at a preset face placing position, acquiring a 3D face image based on depth information aiming at the face of the user;
performing coordinate conversion on the 3D face image to obtain a 3D face point cloud corresponding to the user face;
predicting the fitting parameters required by 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud;
according to the fitting parameters, fitting a preset 3D reference face grid to obtain 3D parametric face data, wherein the 3D reference face grid comprises a plurality of vertexes, the vertexes are in a connection relation, and face semantics are respectively arranged on the vertexes;
and identifying the user according to the 3D parameterized face data.
One or more embodiments of the present specification provide a face recognition apparatus, including:
the system comprises an acquisition module, a display module and a display module, wherein the acquisition module is used for acquiring a 3D face image based on depth information aiming at a user face after the user face is at a preset face placing position;
the coordinate conversion module is used for carrying out coordinate conversion on the 3D face image to obtain a 3D face point cloud corresponding to the user face;
the fitting parameter prediction module predicts fitting parameters required by 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud;
the parameterization module is used for fitting a preset 3D reference face grid according to the fitting parameters to obtain 3D parameterized face data, wherein the 3D reference face grid comprises a plurality of vertexes, the vertexes are in a connection relation, and each vertex is provided with face semantics;
and the identification module is used for identifying the user according to the 3D parameterized face data.
One or more embodiments of the present specification provide a face recognition apparatus, including:
at least one processor; and the number of the first and second groups,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to:
after the face of a user is at a preset face placing position, acquiring a 3D face image based on depth information aiming at the face of the user;
performing coordinate conversion on the 3D face image to obtain a 3D face point cloud corresponding to the user face;
predicting the fitting parameters required by 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud;
according to the fitting parameters, fitting a preset 3D reference face grid to obtain 3D parametric face data, wherein the 3D reference face grid comprises a plurality of vertexes, the vertexes are in a connection relation, and face semantics are respectively arranged on the vertexes;
and identifying the user according to the 3D parameterized face data.
One or more embodiments of the present specification provide a non-transitory computer storage medium storing computer-executable instructions configured to:
after the face of a user is at a preset face placing position, acquiring a 3D face image based on depth information aiming at the face of the user;
performing coordinate conversion on the 3D face image to obtain a 3D face point cloud corresponding to the user face;
predicting the fitting parameters required by 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud;
according to the fitting parameters, fitting a preset 3D reference face grid to obtain 3D parametric face data, wherein the 3D reference face grid comprises a plurality of vertexes, the vertexes are in a connection relation, and face semantics are respectively arranged on the vertexes;
and identifying the user according to the 3D parameterized face data.
At least one technical scheme adopted by one or more embodiments of the specification can achieve the following beneficial effects:
the face recognition is carried out on the user through the 3D face image, compared with a 2D face image, the privacy protection of the user is very friendly, and the privacy safety requirement of the user can be guaranteed. The parameterization of the 3D face image of the user is realized by combining the predicted fitting parameters through the pre-parameterized 3D reference face mesh, the influence of instability of a 3D camera on a face recognition algorithm can be reduced, and the parameterized face data is subjected to parameterization processing, so that the parameterized 3D face data is directly stored without storing a characteristic abstract of data for a bottoming system, and the storage space of the system is effectively reduced.
Drawings
In order to more clearly illustrate the embodiments of the present specification or the technical solutions in the prior art, the drawings needed to be used in the description of the embodiments or the prior art will be briefly introduced below, it is obvious that the drawings in the following description are only some embodiments described in the present specification, and for those skilled in the art, other drawings can be obtained according to the drawings without any creative effort.
Fig. 1 is a schematic flow chart of a face recognition method according to one or more embodiments of the present disclosure;
fig. 2 is a schematic diagram of a 2D face image in an application scenario according to one or more embodiments of the present disclosure;
fig. 3 is a schematic diagram of a 3D face image in an application scenario according to one or more embodiments of the present disclosure;
fig. 4 is a 3D face image with poor imaging effect in an application scenario, which is provided in one or more embodiments of the present disclosure;
fig. 5 is a schematic diagram of coordinate transformation in an application scenario according to one or more embodiments of the present disclosure;
fig. 6 is a schematic diagram of target detection and keypoint detection in an application scenario according to one or more embodiments of the present disclosure;
fig. 7 is a schematic diagram of an implementation process of3D parameterized face data in an application scenario according to one or more embodiments of the present disclosure;
fig. 8 is a schematic diagram of3D parameterized face data in an application scenario according to one or more embodiments of the present disclosure;
fig. 9 is a schematic diagram of face similarity determination in an application scenario according to one or more embodiments of the present disclosure;
fig. 10 is a schematic flowchart of a face recognition method in an application scenario according to one or more embodiments of the present disclosure;
fig. 11 is a schematic structural diagram of a face recognition apparatus according to one or more embodiments of the present disclosure;
fig. 12 is a schematic structural diagram of a face recognition device according to one or more embodiments of the present disclosure.
Detailed Description
The embodiment of the specification provides a face recognition method, a face recognition device, face recognition equipment and a storage medium.
In order to make those skilled in the art better understand the technical solutions in the present specification, the technical solutions in the embodiments of the present specification will be clearly and completely described below with reference to the drawings in the embodiments of the present specification, and it is obvious that the described embodiments are only a part of the embodiments of the present application, and not all of the embodiments. All other embodiments, which can be obtained by a person skilled in the art without making any inventive step based on the embodiments of the present disclosure, shall fall within the scope of protection of the present application.
Fig. 1 is a schematic flowchart of a face recognition method according to one or more embodiments of the present disclosure. The method can be applied to different business fields, such as the field of offline automatic payment, the field of internet financial business, the field of electric business, the field of instant messaging business, the field of game business, the field of official business and the like. The process may be performed by computing devices in the respective field (e.g., face-brushing payment machines for offline payments, etc.), with certain input parameters or intermediate results in the process allowing for manual intervention adjustments to help improve accuracy.
In order to be able to explain the solution shown in fig. 1, a description of face recognition in a conventional solution is given here. Fig. 2 and fig. 3 are schematic diagrams of a 2D face image and a 3D face image in an application scenario, which are provided by one or more embodiments of the present disclosure, respectively, and it can be seen from the diagrams that, compared with the 2D face image, the 3D face image is lower in visibility with naked eyes, and 3D face recognition has an obvious advantage in privacy protection compared with 2D face recognition.
For 3D face recognition, the 3D camera carried by most face brushing machines is a structured light 3D camera with relatively low cost. Such cameras typically acquire a 3D face map based on depth information, the 3D face map storing face data in the form of a 2D map, and for each pixel in the 2D map its corresponding depth information. In the traditional 3D face recognition, the face identity is recognized based on the algorithm analysis of the 3D face image. The defects of the traditional scheme are that the imaging of the structured light 3D camera is unstable, the quality of the 3D face image is easily influenced by objective factors such as environmental factors and distance over-range, and the stability of the 3D face recognition scheme is further influenced. Fig. 4 is a 3D face image with poor imaging effect in an application scenario, which is provided in one or more embodiments of the present disclosure. As can be seen from fig. 4, due to environmental factors, distance over-range, and the like, the quality of the finally acquired 3D face image is poor, and accurate face recognition is difficult to perform.
Based on this, the flow in fig. 1 may include the following steps:
s102: and after the face of the user is at a preset face placing position, acquiring a 3D face image based on depth information aiming at the face of the user.
In the process of executing a service (for example, face-brushing payment and face-brushing access control), a user triggers a face recognition process and needs to acquire a face image of the user. The user can be at general equipment such as smart mobile phone, personal PC, or trigger face identification on dedicated face equipment of brushing (for example, brush face vending machine, brush face machine of ordering by oneself, brush face entrance guard's equipment), and face locating place usually refers to in the preset area in equipment camera the place ahead, and wherein, the position of special equipment in the space of locating is relatively fixed, so its face locating place that corresponds is also relatively fixed. On the device, in order to facilitate the user to find the face placement position quickly, the face placement position can be highlighted for the user in a display form such as a square frame, a circle and the like.
The 3D face map may be a stereoscopic face map generated in three-dimensional space that is capable of storing and presenting to the user the specific shape of the face map in three-dimensional space. However, in practical applications, if the 3D face image is in this form, a large amount of memory space and computing resources are occupied. Therefore, the 3D face image is collected and stored in a 2D image form carrying depth information, and at the moment, the requirements for identification and calculation of the 3D face image can be met, and the occupied storage space is less.
Although the structured light 3D camera is explained as an example in this document, in practical application, the face recognition method mentioned in the embodiments of the present specification can be used for any type of3D camera, and is not limited to the structured light 3D camera or the 3D camera with poor imaging effect. If the imaging effect of the 3D face image acquired by the 3D camera is poor, the accuracy of face recognition can be naturally improved by the face recognition method mentioned in the embodiment of the specification. If the imaging effect of the 3D face image acquired by the 3D camera is better, the accuracy and reliability of the face recognition can be further improved by the face recognition method mentioned in the embodiment of the specification.
S104: and carrying out coordinate conversion on the 3D face image to obtain a 3D face point cloud corresponding to the user face.
Since the 3D face image is acquired based on depth information (for example, when the 3D face image is a 2D face depth image), it needs to be converted into a 3D face point cloud, so as to process the 3D face image in the following.
Specifically, fig. 5 is a schematic diagram of coordinate transformation in an application scenario, provided in one or more embodiments of the present specification. The 2D face depth map collected through the 3D camera usually carries large noise, and at the moment, the 2D face depth map is subjected to noise reduction processing, for example, the 2D face depth map is subjected to noise reduction processing through bilateral filtering. The Bilateral filtering (Bilateral filter) is a nonlinear filtering method, and is a compromise process combining the spatial proximity and the pixel value similarity of an image, and simultaneously considers the spatial information and the gray level similarity to achieve the purpose of edge-preserving and noise-reducing. And performing coordinate conversion according to the depth information corresponding to each pixel in the 2D face depth image after noise reduction so as to convert each pixel from a 2D form pixel coordinate system to a 3D space coordinate system and obtain a 3D point cloud corresponding to the user.
In the process of acquiring the 3D face image, due to the standing position of the user, the acquired 3D face image may include other parts of the user, such as the upper body of the user in the 3D face image shown in fig. 5. At this time, the converted 3D point cloud also includes a point cloud corresponding to the upper body of the user. Based on this, fig. 6 is a schematic diagram of target detection and key point detection in an application scenario provided by one or more embodiments of the present specification, where target detection is performed in a 3D point cloud through a pre-trained 3D target detection model, and a 3D face point cloud corresponding to a user is extracted. The 3D object detection model may be based on deep learning training, such as a convolutional neural network.
S106: and predicting the fitting parameters required by 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud.
In the conventional scheme, 3D face reconstruction can be performed in other ways. For example, a certain number of3D faces are collected in advance, the average shape and average texture part of the faces are obtained, and then the feature vectors of the covariance matrix arranged in descending order according to the feature values are obtained to represent the shape and texture information of the faces, so that a 3D model can be generated in advance. After the 2D face image of the user is obtained, rendering parameters are obtained according to the 2D face image, the rendering parameters and the pre-generated 3D model are optimized together, the finally generated 3D image is enabled to be close to the 2D face image of the user as much as possible, and the purpose of approximately recovering 3D face information from the 2D face image is achieved. However, the scheme is very dependent on the quality of the 2D face image itself, and when the quality of the 2D face image is poor, the recovered 2D face information is naturally also poor.
Fig. 7 is a schematic diagram of an implementation process of3D parameterized face data in an application scenario according to one or more embodiments of the present specification, where in the scheme, parameter prediction is no longer performed through a 2D face image, but a 3D face point cloud is used as an input, and a fitting parameter required for 3D face reconstruction is predicted through a pre-trained fitting parameter prediction model. Compared with a 2D face image, the method has the advantages that the data contained in the 3D face point cloud are more comprehensive and more accurate, and therefore the accuracy of the predicted fitting parameters can be improved. The fitting parameter prediction can be obtained by utilizing a deep learning training neural network model.
S108: and fitting a preset 3D reference face grid according to the fitting parameters to obtain 3D parametric face data, wherein the 3D reference face grid comprises a plurality of vertexes, the vertexes are in a connection relation, and face semantics are respectively arranged on the vertexes.
Fig. 7 is a schematic diagram of an implementation process of3D parameterized face data in an application scenario, where a 3D reference face mesh is pre-constructed, the 3D reference face mesh is different from a 3D model in a conventional scheme, and includes multiple vertices, the multiple vertices have a connection relationship, and each vertex is provided with a face semantic, at this time, it may be considered that the 3D reference face mesh has parameterized each vertex included therein, and thus, after obtaining fitting parameters, fitting processing can be performed based on the fitting parameters, so as to obtain the 3D parameterized face data. For example, the fitting parameter may be α as shown in FIG. 71、α2……αNAfter obtaining the fitting parameters, pass
And performing fitting treatment. Wherein S isnewRepresenting parametric 3D face data,
representing the user's original 3D face point cloud, siIs a representation of the shape of a human face in different dimensions, alphaiIs s isiThe corresponding coefficients.
Parameterization can be understood in the scheme that corresponding parameterization rules (such as the number of vertexes, the connection relation between the vertexes, the face semantics of the vertexes, the position coordinates of key points, the distances between edge points and other points and the like) are set for each vertex in the 3D reference face mesh, the fitting parameters are used as input parameters, the set parameterization rules are not changed for fitting the 3D reference face mesh, and therefore the output parameterized 3D face data still conform to the parameterization rules, which is difficult to achieve in the traditional scheme.
Specifically, the explanation is given by taking a part of rules in parameterized rules as an example. In the 3D reference face mesh, a preset number of vertices (for example, 15000 vertices are included) are set, and each vertex is assigned with a definite face semantic (for example, the vertex No. 5000 represents a nose tip point in a face), and each vertex has a fixed first connection topological relation, and the connection relation between the vertices remains unchanged.
After the fitting parameters are obtained, the position coordinates of at least part of vertexes in the 3D reference face mesh are adjusted on the basis of keeping the first connection topological relation according to the weight of the base vector in the fitting parameters. Therefore, in the obtained 3D parameterized face data, the total number of the vertexes, the first connection topological relation among the vertexes and the face semantics represented by the vertexes do not change, and the parameterization of the 3D face image is realized through the fitting parameters and the 3D reference face mesh. Fig. 8 is a schematic diagram of3D parameterized face data in an application scenario, which is provided in one or more embodiments of the present specification, and it can be clearly observed from the diagram that the detail definition of a 3D face is better than that of an original 3D face point cloud, so as to increase the accuracy of face recognition.
S110: and identifying the user according to the 3D parameterized face data.
A user shoots a face photo of the user through a corresponding terminal in advance, the face photo is used as 3D reserved face data, and the 3D reserved face data is stored in a 3D face database. Generally speaking, a user takes a picture of a face through a 3D camera, and performs parameterization processing on the picture in the above manner, and the processing result is used as 3D set-aside face data. However, when some users leave the end by using a terminal such as a smartphone, it may be difficult to take a 3D face picture due to the camera. At this time, the prediction of the fitting parameters can be performed on the shot 2D face picture, so as to obtain 3D parametric face data as 3D background-left face data. The fitting parameters are predicted through the 2D face picture, and the method can be realized by adopting a training neural network model similarly to the above.
In one or more embodiments of the present specification, the 3D parameterized face data is compared with each 3D background left face data in the 3D face database, the 3D background left face data with the highest similarity is selected, and the user is identified according to the user information corresponding to the 3D background left face data.
Specifically, the 3D face database contains 3D background-left face data corresponding to each user, and when face recognition is performed, the 3D parameterized face data is compared with each 3D background-left face data, so that the similarity between the two can be determined, and the 3D background-left face data with the highest similarity is used as the recognition result of the user, so as to verify the identity information of the user. Fig. 9 is a schematic diagram of face similarity determination in an application scenario according to one or more embodiments of the present disclosure, where in 3D parameterized face data and 3D left-end face data, each vertex has a corresponding position coordinate (x, y, z), where a euclidean distance between corresponding vertices (in 3D parameterized face data, each vertex has a corresponding face semantic, and two vertices having the same face semantic in 3D parameterized face data and 3D left-end face data are considered as corresponding vertices) is calculated, and a smaller euclidean distance indicates that positions of corresponding vertices in a mesh are closer, and a similarity naturally increases. After the Euclidean distances between all the corresponding vertexes are obtained, the similarity between the two vertexes can be obtained through modes of accumulation, mean value calculation and the like. For example, as shown in fig. 9, 3D cut-to-bottom face data corresponding to 3 users (user 1, user 2, and user 3, respectively) are provided, and euclidean distances corresponding to the 3D cut-to-bottom face data are 0.02, 0.4, and 0.3, respectively, where the cut-to-bottom of user 1 with the smallest euclidean distance is selected as the 3D cut-to-bottom face data with the highest similarity, and face recognition is performed. Of course, as already explained above, the target detection is performed in the 3D point cloud of the user, resulting in a 3D face point cloud. Generally speaking, in the target detection process, the acquired 3D face point clouds are actually aligned, so that the euclidean distance between corresponding vertices can be directly obtained to represent the similarity during similarity comparison.
Further, in determining the similarity, if the 3D background human face data has an earlier background time, the face shape of the user may have a larger change for some objective reasons. At this time, if the euclidean distance calculated by accumulation, averaging, etc. is used, it may not be accurate any more.
Based on this, the 3D parameterized face data (including the 3D left-end face data, of course) is pre-divided into a plurality of face regions, each region is provided with a fixed vertex, for example, 0-999 is a forehead region, 1000-. The face variation level refers to the variation degree of each region with time under the influence of natural physiological factors (excluding artificial factors such as operation and injury). The higher the face change level, the faster the change degree is indicated. Generally, the edge area has the highest level of variation corresponding to the human face because of no skeletal support.
At this time, each face region in the 3D parameterized face data is compared with the corresponding face region in the 3D background-left face data to obtain a first sub-similarity for the face region. The first sub-similarity may be determined based on the euclidean distance between vertices, similar to the similarity above. At this time, on the basis of the first sub-similarity, the influence of time on the face shape is considered, and the first sub-similarity is compensated according to the bottom-left time of the 3D bottom-left face data and the face change level of the face area, so that a second sub-similarity is obtained. The longer the margin time is, the higher the face change level is, the larger the change degree of the face shape is, and the higher the compensation degree is, and the positive correlation is formed between the compensation degree and the margin time and the face change level. After the second sub-similarity of all the face regions is obtained, the similarity between the 3D parametric face data and the 3D background-left face data can be obtained through modes of summation, averaging and the like, so that the influence of time on the face shape can be combined, and the face recognition of the user with long background-left time can be accurately carried out.
In one or more embodiments of the present disclosure, as described above, the target detection is performed in the 3D point cloud of the user, the 3D face point cloud of the user is identified, and the prediction of the fitting parameter is performed based on the 3D face point cloud. In order to increase the prediction accuracy of the fitting parameters and the smoothness of the 3D reference face mesh during the subsequent fitting process, after the 3D face point cloud of the user is obtained, the 3D face point cloud may be subjected to key point detection by using a pre-trained face key point detection model (for example, the pre-trained face key point detection model may be similar to the 3D target detection model and generated by convolutional neural network training), so as to extract key point information of the user face. For example, as shown in fig. 6, the key point information may include a nose tip, a mouth corner, a pupil, and the like.
After the key point information is extracted, the fitting parameters are predicted through the key point information and the 3D face point cloud, so that the prediction accuracy of the fitting parameters is improved. As shown in fig. 7, when fitting the 3D reference face mesh, the key point information may be fitted first, and adaptive smoothing may be performed based on the key point information in the fitting process of other vertices, so as to increase the smoothness of the 3D parametric face data.
Further, the 3D face data is set as a base, which can improve the privacy security of the user, but it is still difficult to meet the requirements for some scenes with higher security level requirements, and the storage space and the computing resources used by the scenes are also larger.
Based on the above, after the corresponding key point information in the 3D parameterized face data and the 3D left-bottom face data is obtained, a second connection topological relation is generated according to the position coordinates of each key point, and the second connection topological relation only contains the key point information and does not contain information of other vertices. The key point information is a vertex of a core part in the face, can reflect basic information of the face, and when determining the similarity between the 3D parameterized face data and each 3D bottom-left face data through the second connection topological relation (for example, determining the similarity through the euclidean distance between the two, the calculation speed when the two are attached and compared, the calculation resource consumption and the like), on the premise of consuming less calculation resources, a more accurate recognition effect can be realized, so that the specified 3D bottom-left face data with the similarity exceeding a preset threshold is selected, and the 3D face recognition is performed on the user.
At this time, the security level corresponding to the specified 3D copy face data is determined, and the security level may be preset based on user requirements, business requirements, and the like. When the security level is lower (lower than the preset level), the designated 3D background-preserving face data obtained by screening the second connection topological relation can be considered as a primary screening process, and secondary screening is performed on the designated 3D background-preserving face data through the more detailed first connection topological relation, so that the designated 3D background-preserving face data with the highest similarity is selected and used for performing 3D face recognition on the user, and therefore large-area primary screening is performed through the second connection topological relation with less computing resource consumption, and the consumption of the whole computing resources is reduced. When the security level is higher (higher than the preset level), deleting the first connection topological relation in the specified 3D copy face in the 3D face database, replacing the second connection topological relation, and storing the second connection topological relation in the 3D copy face data.
Fig. 10 is a flowchart of a face recognition method in an application scenario, according to one or more embodiments of the present disclosure. When the scheme is applied to online face brushing payment, firstly, in a 3D face data acquisition and preprocessing module, a 2D face depth map of a user is acquired through a structured light 3D camera, preprocessing (such as noise reduction processing and coordinate conversion) is carried out on the map, and 3D face information (including 3D face point cloud and key point information) is obtained through face target recognition. And then, in a 3D face parameterization module, carrying out 3D face parameterization through a preset 3D reference face grid to obtain 3D parameterized face data. And finally, determining 3D background face data in a 3D face database in a parameterized 3D face recognition module, and finally determining user information by performing comparison retrieval.
On the basis of the same idea, one or more embodiments of the present specification further provide apparatuses and devices corresponding to the above-described method, as shown in fig. 11 and 12.
Fig. 11 is a schematic structural diagram of a face recognition apparatus according to one or more embodiments of the present disclosure, where the apparatus includes:
the acquisition module 1102 is used for acquiring a 3D face image based on depth information aiming at a user face after the user face is at a preset face placing position;
a coordinate conversion module 1104, configured to perform coordinate conversion on the 3D face image to obtain a 3D face point cloud corresponding to the user face;
a fitting parameter prediction module 1106 for predicting the fitting parameters required by 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud;
a parameterization module 1108, configured to perform fitting processing on a preset 3D reference face mesh according to the fitting parameters to obtain 3D parameterized face data, where the 3D reference face mesh includes multiple vertices, the multiple vertices have a connection relationship, and each vertex is provided with a face semantic;
the recognition module 1110 recognizes the user according to the 3D parameterized face data.
Optionally, the number of vertices in the 3D reference face mesh is fixed, and the connection relationship among the multiple vertices is a fixed first connection topological relationship;
the parameterization module 1108 adjusts the position coordinates of at least part of the vertices in the 3D reference face mesh based on the first connection topology relationship according to the basis vector weights in the fitting parameters, so as to obtain 3D parameterized face data.
Optionally, the 3D face image is a 2D face depth image;
the coordinate conversion module 1104 is used for performing noise reduction processing on the 2D face depth map;
performing coordinate conversion according to the depth information corresponding to each pixel in the 2D face depth image after noise reduction to convert each pixel from a 2D pixel coordinate system to a 3D space coordinate system to obtain a 3D point cloud corresponding to the user;
and carrying out target detection on the 3D point cloud through a pre-trained 3D target detection model so as to determine a 3D face point cloud corresponding to the user.
Optionally, the method further comprises:
a key point identification module 1112, configured to perform key point detection on the 3D face point cloud through a pre-trained face key point detection model to determine key point information of the user face therein;
the fitting parameter prediction module 1106 predicts the fitting parameters required for 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud and the key point information.
Optionally, the identifying module 1110 compares the 3D parameterized face data with each 3D left-end face data in a 3D face database to determine a similarity between the 3D parameterized face data and each 3D left-end face data, where the similarity is determined according to an euclidean distance between each corresponding vertex in the 3D parameterized face data and the 3D left-end face data;
and 3D face recognition is carried out on the user according to the 3D reserved face data with the highest similarity.
Optionally, the identification module 1110 determines a plurality of divided face regions in the 3D parameterized face data, and determines a face change level corresponding to each face region;
aiming at each face region in the 3D parametric face data, comparing the face region with the face region corresponding to each 3D bottom-left face data in a 3D face database to obtain a first sub-similarity corresponding to the face region, and compensating the first sub-similarity according to the bottom-left time of each 3D bottom-left face data and the corresponding face change grade to obtain a second sub-similarity, wherein the compensation degree is positively correlated with the bottom-left time and the face change grade;
and according to the second sub-similarity of each face region, obtaining the similarity between the 3D parameterized face data and each 3D background-left face data.
Optionally, the identification module 1110 determines the 3D parameterized face data, each 3D background-left face data in the 3D face database, and corresponding key point information respectively;
generating respective corresponding second connection topological relations according to the key point information;
determining the similarity between the 3D parameterized face data and each 3D set-bottom face data according to the second connection topological relation;
and selecting the specified 3D background left face data with the similarity exceeding a preset threshold value from the 3D face database, so as to perform 3D face recognition on the user through the specified 3D background left face data.
Optionally, connection relations among the plurality of vertices in the 3D reference face mesh are fixed first connection topological relations;
the identification module 1110 determines a security level corresponding to the specified 3D copy-on-the-back face data;
if the security level is lower than a preset level, selecting specified 3D background-preserving face data with the highest similarity from the specified 3D background-preserving face data according to the first connection topological relation so as to perform 3D face recognition on the user;
and if the safety level is higher than the preset level, deleting the first connection topological relation in the specified 3D copy-on-the-bottom face in the 3D face database, and storing the second connection topological relation in the specified 3D copy-on-the-bottom face data as a substitute.
Fig. 12 is a schematic structural diagram of a face recognition device according to one or more embodiments of the present specification, where the face recognition device includes:
at least one processor; and the number of the first and second groups,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to:
after the face of a user is at a preset face placing position, acquiring a 3D face image based on depth information aiming at the face of the user;
performing coordinate conversion on the 3D face image to obtain a 3D face point cloud corresponding to the user face;
predicting the fitting parameters required by 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud;
according to the fitting parameters, fitting a preset 3D reference face grid to obtain 3D parametric face data, wherein the 3D reference face grid comprises a plurality of vertexes, the vertexes are in a connection relation, and face semantics are respectively arranged on the vertexes;
and identifying the user according to the 3D parameterized face data.
Based on the same idea, one or more embodiments of the present specification further provide a non-volatile computer storage medium corresponding to the above method, and storing computer-executable instructions configured to:
after the face of a user is at a preset face placing position, acquiring a 3D face image based on depth information aiming at the face of the user;
performing coordinate conversion on the 3D face image to obtain a 3D face point cloud corresponding to the user face;
predicting the fitting parameters required by 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud;
according to the fitting parameters, fitting a preset 3D reference face grid to obtain 3D parametric face data, wherein the 3D reference face grid comprises a plurality of vertexes, the vertexes are in a connection relation, and face semantics are respectively arranged on the vertexes;
and identifying the user according to the 3D parameterized face data.
In the 90 s of the 20 th century, improvements in a technology could clearly distinguish between improvements in hardware (e.g., improvements in circuit structures such as diodes, transistors, switches, etc.) and improvements in software (improvements in process flow). However, as technology advances, many of today's process flow improvements have been seen as direct improvements in hardware circuit architecture. Designers almost always obtain the corresponding hardware circuit structure by programming an improved method flow into the hardware circuit. Thus, it cannot be said that an improvement in the process flow cannot be realized by hardware physical modules. For example, a Programmable Logic Device (PLD), such as a Field Programmable Gate Array (FPGA), is an integrated circuit whose Logic functions are determined by programming the Device by a user. A digital system is "integrated" on a PLD by the designer's own programming without requiring the chip manufacturer to design and fabricate application-specific integrated circuit chips. Furthermore, nowadays, instead of manually making an Integrated Circuit chip, such Programming is often implemented by "logic compiler" software, which is similar to a software compiler used in program development and writing, but the original code before compiling is also written by a specific Programming Language, which is called Hardware Description Language (HDL), and HDL is not only one but many, such as abel (advanced Boolean Expression Language), ahdl (alternate Hardware Description Language), traffic, pl (core universal Programming Language), HDCal (jhdware Description Language), lang, Lola, HDL, laspam, hardward Description Language (vhr Description Language), vhal (Hardware Description Language), and vhigh-Language, which are currently used in most common. It will also be apparent to those skilled in the art that hardware circuitry that implements the logical method flows can be readily obtained by merely slightly programming the method flows into an integrated circuit using the hardware description languages described above.
The controller may be implemented in any suitable manner, for example, the controller may take the form of, for example, a microprocessor or processor and a computer-readable medium storing computer-readable program code (e.g., software or firmware) executable by the (micro) processor, logic gates, switches, an Application Specific Integrated Circuit (ASIC), a programmable logic controller, and an embedded microcontroller, examples of which include, but are not limited to, the following microcontrollers: the ARC625D, Atmel AT91SAM, Microchip PIC18F26K20, and Silicone Labs C8051F320, the memory controller may also be implemented as part of the control logic for the memory. Those skilled in the art will also appreciate that, in addition to implementing the controller as pure computer readable program code, the same functionality can be implemented by logically programming method steps such that the controller is in the form of logic gates, switches, application specific integrated circuits, programmable logic controllers, embedded microcontrollers and the like. Such a controller may thus be considered a hardware component, and the means included therein for performing the various functions may also be considered as a structure within the hardware component. Or even means for performing the functions may be regarded as being both a software module for performing the method and a structure within a hardware component.
The systems, devices, modules or units illustrated in the above embodiments may be implemented by a computer chip or an entity, or by a product with certain functions. One typical implementation device is a computer. In particular, the computer may be, for example, a personal computer, a laptop computer, a cellular telephone, a camera phone, a smartphone, a personal digital assistant, a media player, a navigation device, an email device, a game console, a tablet computer, a wearable device, or a combination of any of these devices.
For convenience of description, the above devices are described as being divided into various units by function, and are described separately. Of course, the functions of the various elements may be implemented in the same one or more software and/or hardware implementations of the present description.
As will be appreciated by one skilled in the art, the present specification embodiments may be provided as a method, system, or computer program product. Accordingly, embodiments of the present description may take the form of an entirely hardware embodiment, an entirely software embodiment or an embodiment combining software and hardware aspects. Furthermore, embodiments of the present description may take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, and so forth) having computer-usable program code embodied therein.
The description has been presented with reference to flowchart illustrations and/or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the description. It will be understood that each flow and/or block of the flow diagrams and/or block diagrams, and combinations of flows and/or blocks in the flow diagrams and/or block diagrams, can be implemented by computer program instructions. These computer program instructions may be provided to a processor of a general purpose computer, special purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create means for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be stored in a computer-readable memory that can direct a computer or other programmable data processing apparatus to function in a particular manner, such that the instructions stored in the computer-readable memory produce an article of manufacture including instruction means which implement the function specified in the flowchart flow or flows and/or block diagram block or blocks.
These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer implemented process such that the instructions which execute on the computer or other programmable apparatus provide steps for implementing the functions specified in the flowchart flow or flows and/or block diagram block or blocks.
In a typical configuration, a computing device includes one or more processors (CPUs), input/output interfaces, network interfaces, and memory.
The memory may include forms of volatile memory in a computer readable medium, Random Access Memory (RAM) and/or non-volatile memory, such as Read Only Memory (ROM) or flash memory (flash RAM). Memory is an example of a computer-readable medium.
Computer-readable media, including both non-transitory and non-transitory, removable and non-removable media, may implement information storage by any method or technology. The information may be computer readable instructions, data structures, modules of a program, or other data. Examples of computer storage media include, but are not limited to, phase change memory (PRAM), Static Random Access Memory (SRAM), Dynamic Random Access Memory (DRAM), other types of Random Access Memory (RAM), Read Only Memory (ROM), Electrically Erasable Programmable Read Only Memory (EEPROM), flash memory or other memory technology, compact disc read only memory (CD-ROM), Digital Versatile Discs (DVD) or other optical storage, magnetic cassettes, magnetic tape magnetic disk storage or other magnetic storage devices, or any other non-transmission medium that can be used to store information that can be accessed by a computing device. As defined herein, a computer readable medium does not include a transitory computer readable medium such as a modulated data signal and a carrier wave.
It should also be noted that the terms "comprises," "comprising," or any other variation thereof, are intended to cover a non-exclusive inclusion, such that a process, method, article, or apparatus that comprises a list of elements does not include only those elements but may include other elements not expressly listed or inherent to such process, method, article, or apparatus. Without further limitation, an element defined by the phrase "comprising an … …" does not exclude the presence of other like elements in a process, method, article, or apparatus that comprises the element.
This description may be described in the general context of computer-executable instructions, such as program modules, being executed by a computer. Generally, program modules include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types. The specification may also be practiced in distributed computing environments where tasks are performed by remote processing devices that are linked through a communications network. In a distributed computing environment, program modules may be located in both local and remote computer storage media including memory storage devices.
The embodiments in the present specification are described in a progressive manner, and the same and similar parts among the embodiments are referred to each other, and each embodiment focuses on the differences from the other embodiments. In particular, for the embodiments of the apparatus, the device, and the nonvolatile computer storage medium, since they are substantially similar to the embodiments of the method, the description is simple, and for the relevant points, reference may be made to the partial description of the embodiments of the method.
The foregoing description has been directed to specific embodiments of this disclosure. Other embodiments are within the scope of the following claims. In some cases, the actions or steps recited in the claims may be performed in a different order than in the embodiments and still achieve desirable results. In addition, the processes depicted in the accompanying figures do not necessarily require the particular order shown, or sequential order, to achieve desirable results. In some embodiments, multitasking and parallel processing may also be possible or may be advantageous.
The above description is merely one or more embodiments of the present disclosure and is not intended to limit the present disclosure. Various modifications and alterations to one or more embodiments of the present description will be apparent to those skilled in the art. Any modification, equivalent replacement, improvement or the like made within the spirit and principle of one or more embodiments of the present specification should be included in the scope of the claims of the present specification.
Claims (17)
1. A face recognition method, comprising:
after the face of a user is at a preset face placing position, acquiring a 3D face image based on depth information aiming at the face of the user;
performing coordinate conversion on the 3D face image to obtain a 3D face point cloud corresponding to the user face;
predicting the fitting parameters required by 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud;
according to the fitting parameters, fitting a preset 3D reference face grid to obtain 3D parametric face data, wherein the 3D reference face grid comprises a plurality of vertexes, the vertexes are in a connection relation, and face semantics are respectively arranged on the vertexes;
and identifying the user according to the 3D parameterized face data.
2. The method of claim 1, wherein the number of vertices in the 3D reference face mesh is fixed, and the connection relationship between the vertices is a fixed first connection topology relationship;
according to the fitting parameters, fitting a preset 3D reference face grid to obtain 3D parametric face data, and specifically comprises the following steps:
and adjusting the position coordinates of at least part of vertexes in the 3D reference face mesh on the basis of keeping the first connection topological relation according to the weight of the base vector in the fitting parameters to obtain 3D parametric face data.
3. The method of claim 1, wherein the 3D face map is a 2D face depth map;
the coordinate conversion is performed on the 3D face image to obtain a 3D face point cloud corresponding to the user face, and the method specifically includes:
carrying out noise reduction processing on the 2D face depth image;
performing coordinate conversion according to the depth information corresponding to each pixel in the 2D face depth image after noise reduction to convert each pixel from a 2D pixel coordinate system to a 3D space coordinate system to obtain a 3D point cloud corresponding to the user;
and carrying out target detection on the 3D point cloud through a pre-trained 3D target detection model so as to determine a 3D face point cloud corresponding to the user.
4. The method of claim 3, the target detecting the 3D point cloud by a pre-trained 3D target detection model to, after determining therein a corresponding 3D face point cloud for the user, further comprising:
performing key point detection on the 3D face point cloud through a pre-trained face key point detection model to determine key point information of the user face;
the method for predicting the fitting parameters required by 3D face reconstruction according to the 3D face point cloud and through a pre-trained fitting parameter prediction model specifically comprises the following steps:
and predicting the fitting parameters required by 3D face reconstruction according to the 3D face point cloud and the key point information through a pre-trained fitting parameter prediction model.
5. The method according to claim 1, wherein the identifying the user based on the 3D parameterized face data comprises:
comparing the 3D parameterized face data with each 3D left-end face data in a 3D face database to determine the similarity between the 3D parameterized face data and each 3D left-end face data, wherein the similarity is determined according to Euclidean distances between corresponding vertexes in the 3D parameterized face data and the 3D left-end face data;
and 3D face recognition is carried out on the user according to the 3D reserved face data with the highest similarity.
6. The method of claim 5, wherein comparing the 3D parameterized face data with each 3D background face data in a 3D face database to determine the similarity between the 3D parameterized face data and each of the 3D background face data, comprises:
determining a plurality of divided face regions in the 3D parametric face data, and determining face change levels corresponding to the face regions;
aiming at each face region in the 3D parametric face data, comparing the face region with the face region corresponding to each 3D bottom-left face data in a 3D face database to obtain a first sub-similarity corresponding to the face region, and compensating the first sub-similarity according to the bottom-left time of each 3D bottom-left face data and the corresponding face change grade to obtain a second sub-similarity, wherein the compensation degree is positively correlated with the bottom-left time and the face change grade;
and according to the second sub-similarity of each face region, obtaining the similarity between the 3D parameterized face data and each 3D background-left face data.
7. The method according to claim 4, wherein the identifying the user based on the 3D parameterized face data comprises:
determining key point information corresponding to the 3D parameterized face data and each 3D reserved-base face data in the 3D face database respectively;
generating respective corresponding second connection topological relations according to the key point information;
determining the similarity between the 3D parameterized face data and each 3D set-bottom face data according to the second connection topological relation;
and selecting the specified 3D background left face data with the similarity exceeding a preset threshold value from the 3D face database, so as to perform 3D face recognition on the user through the specified 3D background left face data.
8. The method of claim 7, wherein the connection relationships between the plurality of vertices in the 3D reference face mesh are fixed first connection topological relationships;
the 3D face recognition of the user through the specified 3D background left face data specifically includes:
determining a safety level corresponding to the specified 3D background left face data;
if the security level is lower than a preset level, selecting specified 3D background-preserving face data with the highest similarity from the specified 3D background-preserving face data according to the first connection topological relation so as to perform 3D face recognition on the user;
and if the safety level is higher than the preset level, deleting the first connection topological relation in the specified 3D copy-on-the-bottom face in the 3D face database, and storing the second connection topological relation in the specified 3D copy-on-the-bottom face data as a substitute.
9. A face recognition apparatus comprising:
the system comprises an acquisition module, a display module and a display module, wherein the acquisition module is used for acquiring a 3D face image based on depth information aiming at a user face after the user face is at a preset face placing position;
the coordinate conversion module is used for carrying out coordinate conversion on the 3D face image to obtain a 3D face point cloud corresponding to the user face;
the fitting parameter prediction module predicts fitting parameters required by 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud;
the parameterization module is used for fitting a preset 3D reference face grid according to the fitting parameters to obtain 3D parameterized face data, wherein the 3D reference face grid comprises a plurality of vertexes, the vertexes are in a connection relation, and each vertex is provided with face semantics;
and the identification module is used for identifying the user according to the 3D parameterized face data.
10. The apparatus of claim 9, wherein the number of vertices in the 3D reference face mesh is fixed, and the connection relationship between the vertices is a fixed first connection topology relationship;
and the parameterization module adjusts the position coordinates of at least part of vertexes in the 3D reference face mesh on the basis of keeping the first connection topological relation according to the weight of the base vector in the fitting parameter, so as to obtain 3D parameterized face data.
11. The apparatus of claim 9, the 3D face map is a 2D face depth map;
the coordinate conversion module is used for carrying out noise reduction processing on the 2D face depth map;
performing coordinate conversion according to the depth information corresponding to each pixel in the 2D face depth image after noise reduction to convert each pixel from a 2D pixel coordinate system to a 3D space coordinate system to obtain a 3D point cloud corresponding to the user;
and carrying out target detection on the 3D point cloud through a pre-trained 3D target detection model so as to determine a 3D face point cloud corresponding to the user.
12. The apparatus of claim 11, further comprising:
the key point identification module is used for detecting key points of the 3D face point cloud through a pre-trained face key point detection model so as to determine key point information of the user face;
and the fitting parameter prediction module predicts the fitting parameters required by 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud and the key point information.
13. The apparatus of claim 9, the recognition module to compare the 3D parameterized face data to each 3D backing face data in a 3D face database to determine a similarity between the 3D parameterized face data and each of the 3D backing face data, the similarity determined based on euclidean distances between corresponding vertices in the 3D parameterized face data and the 3D backing face data;
and 3D face recognition is carried out on the user according to the 3D reserved face data with the highest similarity.
14. The apparatus of claim 13, wherein the recognition module determines a plurality of face regions divided in the 3D parameterized face data and determines a face variation level corresponding to each face region;
aiming at each face region in the 3D parametric face data, comparing the face region with the face region corresponding to each 3D bottom-left face data in a 3D face database to obtain a first sub-similarity corresponding to the face region, and compensating the first sub-similarity according to the bottom-left time of each 3D bottom-left face data and the corresponding face change grade to obtain a second sub-similarity, wherein the compensation degree is positively correlated with the bottom-left time and the face change grade;
and according to the second sub-similarity of each face region, obtaining the similarity between the 3D parameterized face data and each 3D background-left face data.
15. The apparatus of claim 12, wherein the recognition module determines corresponding keypoint information for each of the 3D parameterized face data, the 3D footed face data in the 3D face database;
generating respective corresponding second connection topological relations according to the key point information;
determining the similarity between the 3D parameterized face data and each 3D set-bottom face data according to the second connection topological relation;
and selecting the specified 3D reserved face data with the similarity exceeding a preset threshold value from the 3D face database, so as to perform 3D face recognition on the user through the specified 3D reserved face data.
16. The apparatus of claim 15, wherein connection relationships between the plurality of vertices in the 3D reference face mesh are fixed first connection topological relationships;
the identification module is used for determining the safety level corresponding to the specified 3D background left face data;
if the security level is lower than a preset level, selecting specified 3D background-preserving face data with the highest similarity from the specified 3D background-preserving face data according to the first connection topological relation so as to perform 3D face recognition on the user;
and if the safety level is higher than the preset level, deleting the first connection topological relation in the specified 3D copy-on-the-bottom face in the 3D face database, and storing the second connection topological relation in the specified 3D copy-on-the-bottom face data as a substitute.
17. A face recognition device comprising:
at least one processor; and the number of the first and second groups,
a memory communicatively coupled to the at least one processor; wherein the content of the first and second substances,
the memory stores instructions executable by the at least one processor to enable the at least one processor to:
after the face of a user is at a preset face placing position, acquiring a 3D face image based on depth information aiming at the face of the user;
performing coordinate conversion on the 3D face image to obtain a 3D face point cloud corresponding to the user face;
predicting the fitting parameters required by 3D face reconstruction through a pre-trained fitting parameter prediction model according to the 3D face point cloud;
according to the fitting parameters, fitting a preset 3D reference face grid to obtain 3D parametric face data, wherein the 3D reference face grid comprises a plurality of vertexes, the vertexes are in a connection relation, and face semantics are respectively arranged on the vertexes;
and identifying the user according to the 3D parameterized face data.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210181933.XA CN114511911A (en) | 2022-02-25 | 2022-02-25 | Face recognition method, device and equipment |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202210181933.XA CN114511911A (en) | 2022-02-25 | 2022-02-25 | Face recognition method, device and equipment |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN114511911A true CN114511911A (en) | 2022-05-17 |
Family
ID=81552921
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202210181933.XA Pending CN114511911A (en) | 2022-02-25 | 2022-02-25 | Face recognition method, device and equipment |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN114511911A (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115422863A (en) * | 2022-08-31 | 2022-12-02 | 山东启芯软件科技有限公司 | Embedded logic optimization verification flow framework and verification method |
Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150278582A1 (en) * | 2014-03-27 | 2015-10-01 | Avago Technologies General Ip (Singapore) Pte. Ltd | Image Processor Comprising Face Recognition System with Face Recognition Based on Two-Dimensional Grid Transform |
| WO2018080702A1 (en) * | 2016-10-31 | 2018-05-03 | Google Llc | Face reconstruction from a learned embedding |
| WO2019080488A1 (en) * | 2017-10-27 | 2019-05-02 | 东南大学 | Three-dimensional human face recognition method based on multi-scale covariance descriptor and local sensitive riemann kernel sparse classification |
| CN109978984A (en) * | 2017-12-27 | 2019-07-05 | Tcl集团股份有限公司 | Face three-dimensional rebuilding method and terminal device |
| CN111523398A (en) * | 2020-03-30 | 2020-08-11 | 西安交通大学 | Method and device for fusing 2D face detection and 3D face recognition |
| CN113392763A (en) * | 2021-06-15 | 2021-09-14 | 支付宝(杭州)信息技术有限公司 | Face recognition method, device and equipment |
| CN113632098A (en) * | 2021-07-02 | 2021-11-09 | 华为技术有限公司 | Face image processing method and device and vehicle |
-
2022
- 2022-02-25 CN CN202210181933.XA patent/CN114511911A/en active Pending
Patent Citations (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20150278582A1 (en) * | 2014-03-27 | 2015-10-01 | Avago Technologies General Ip (Singapore) Pte. Ltd | Image Processor Comprising Face Recognition System with Face Recognition Based on Two-Dimensional Grid Transform |
| WO2018080702A1 (en) * | 2016-10-31 | 2018-05-03 | Google Llc | Face reconstruction from a learned embedding |
| WO2019080488A1 (en) * | 2017-10-27 | 2019-05-02 | 东南大学 | Three-dimensional human face recognition method based on multi-scale covariance descriptor and local sensitive riemann kernel sparse classification |
| CN109978984A (en) * | 2017-12-27 | 2019-07-05 | Tcl集团股份有限公司 | Face three-dimensional rebuilding method and terminal device |
| CN111523398A (en) * | 2020-03-30 | 2020-08-11 | 西安交通大学 | Method and device for fusing 2D face detection and 3D face recognition |
| CN113392763A (en) * | 2021-06-15 | 2021-09-14 | 支付宝(杭州)信息技术有限公司 | Face recognition method, device and equipment |
| CN113632098A (en) * | 2021-07-02 | 2021-11-09 | 华为技术有限公司 | Face image processing method and device and vehicle |
Non-Patent Citations (2)
| Title |
|---|
| 李佳;任慧;: "基于Kinect 2的三维人脸个性化建模", 中国传媒大学学报(自然科学版), no. 02, 25 April 2018 (2018-04-25) * |
| 李晔: "非特定条件下三维人脸识别关键技术研究", 中国博士学位论文全文数据库 信息科技辑, no. 01, 15 January 2018 (2018-01-15) * |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN115422863A (en) * | 2022-08-31 | 2022-12-02 | 山东启芯软件科技有限公司 | Embedded logic optimization verification flow framework and verification method |
| CN115422863B (en) * | 2022-08-31 | 2023-10-10 | 山东启芯软件科技有限公司 | Embedded logic optimization verification flow framework and verification method |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN109255352B (en) | Target detection method, device and system | |
| CN110163198B (en) | Table identification reconstruction method and device and storage medium | |
| Xie et al. | Multilevel cloud detection in remote sensing images based on deep learning | |
| Luo et al. | Aircraft target detection in remote sensing images based on improved YOLOv5 | |
| CN113095124B (en) | Face living body detection method and device and electronic equipment | |
| Xu et al. | Effective face detector based on yolov5 and superresolution reconstruction | |
| CN106407911A (en) | Image-based eyeglass recognition method and device | |
| CN113239907B (en) | Face recognition detection method and device, electronic equipment and storage medium | |
| CN110852257B (en) | Method and device for detecting key points of human face and storage medium | |
| CN109711416A (en) | Target identification method, device, computer equipment and storage medium | |
| JP2014164656A (en) | Image processing method and program | |
| Du et al. | Improving RGBD saliency detection using progressive region classification and saliency fusion | |
| Ruppert et al. | Robust fitting of parallax-aware mixtures for path guiding | |
| CN113569598A (en) | Image processing method and image processing apparatus | |
| CN112836625A (en) | Face living body detection method and device and electronic equipment | |
| CN112446322A (en) | Eyeball feature detection method, device, equipment and computer-readable storage medium | |
| Song et al. | Depth-aware saliency detection using discriminative saliency fusion | |
| CN111368865B (en) | Remote sensing image oil storage tank detection method and device, readable storage medium and equipment | |
| CN114511911A (en) | Face recognition method, device and equipment | |
| CN111310531A (en) | Image classification method and device, computer equipment and storage medium | |
| CN112597909A (en) | Method and equipment for evaluating quality of face picture | |
| CN117036829A (en) | Method and system for achieving label enhancement based on prototype learning for identifying fine granularity of blade | |
| CN113658180B (en) | Surface defect region segmentation method and device based on spatial context guidance | |
| CN111753583A (en) | Identification method and device | |
| CN111967365B (en) | Image connection point extraction method and device |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination |