CN109190449A - Age recognition methods, device, electronic equipment and storage medium - Google Patents
Age recognition methods, device, electronic equipment and storage medium Download PDFInfo
- Publication number
- CN109190449A CN109190449A CN201810747040.0A CN201810747040A CN109190449A CN 109190449 A CN109190449 A CN 109190449A CN 201810747040 A CN201810747040 A CN 201810747040A CN 109190449 A CN109190449 A CN 109190449A
- Authority
- CN
- China
- Prior art keywords
- facial image
- age
- face
- estimation
- target
- 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
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/172—Classification, e.g. identification
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/161—Detection; Localisation; Normalisation
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/178—Human faces, e.g. facial parts, sketches or expressions estimating age from face image; using age information for improving recognition
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of age recognition methods, device, electronic equipment and storage mediums, wherein the described method includes: the facial image in detection target video, obtains target face image set;Age identification and quality analysis are carried out to each facial image that the target facial image is concentrated, obtain estimation age and the quality score of each facial image;The estimation age of the affiliated personnel of each facial image is determined according to the estimation age of each facial image and quality score.It solves the problems, such as that age recognition accuracy is relatively low, the influence of picture quality is considered when carrying out age identification, the accuracy of age identification can be improved.
Description
Technical field
This disclosure relates to software technical field of face recognition more particularly to a kind of age recognition methods, device, electronic equipment
And storage medium.
Background technique
Face recognition technology specifically includes that gender identification, age identification, Expression Recognition etc..Wherein, age identification can lead to
Facial image is crossed to identify to obtain the age information of people.
In the related technology, age identification is mainly by judging multiframe picture.It specifically includes that firstly, from video
Several intracoded frames are extracted as key frame;Then, recognition of face is carried out to each key frame and judges the age;Finally, by each
The age of key frame is averaged, and the final age is obtained.
Summary of the invention
It is found in the practical application of above-mentioned the relevant technologies, since frame image quality each in video differs greatly, leads to year
The age accuracy of identification is lower.
To overcome the problems in correlation technique, the disclosure provide a kind of age recognition methods, device, electronic equipment and
Storage medium.
According to the first aspect of the embodiments of the present disclosure, a kind of age recognition methods is provided, comprising:
The facial image in target video is detected, target face image set is obtained;
Age identification and quality analysis are carried out to each facial image that the target facial image is concentrated, obtain each face figure
The estimation age of picture and quality score;
The estimation age of the affiliated personnel of each facial image is determined according to the estimation age of each facial image and quality score.
Optionally, the above-mentioned each facial image concentrated to the target facial image carries out age identification and quality analysis,
Obtain estimation age and the quality score of each facial image, comprising:
Face cluster is carried out to the target face image set, obtains at least one facial image group;
For each facial image group, age identification and quality point are carried out to facial image each in the facial image group respectively
Analysis, obtains estimation age and the quality score of each facial image.
Optionally, above-mentioned to determine the affiliated personnel's of each facial image according to the estimation age of each facial image and quality score
Estimate the age, comprising:
For each facial image group, obtained according to the estimation age of each facial image in the facial image group and quality
Divide the estimation age for determining the facial image group, wherein each facial image in each facial image group belongs to same personnel.
Optionally, above-mentioned for each facial image group, according to the estimation of each facial image in the facial image group
Age and quality score determine the estimation age of the facial image group, comprising:
For each facial image group, by the estimation age of each facial image in the facial image group and quality score
It is multiplied, obtains the weighting age of each facial image;
For each facial image group, the weighting age of each facial image in the facial image group is added, is obtained described
The estimation age of facial image group.
Optionally, above-mentioned target face image set includes first object facial image subset, in the detection target video
Facial image, obtain target face image set, comprising:
Face datection is carried out to the first frame picture of the target video, obtains first with reference to facial image;
Facial image is referred to according to described first, face tracking is carried out to remaining picture of the target video, obtains the
One target facial image subset.
Optionally, above-mentioned target face image set further includes the second target facial image subset, the above method further include:
If in remaining picture of the target video a frame picture carry out face tracking failure, to the picture into
Row Face datection obtains second when detecting face with reference to facial image;
According to described second refer to facial image, to after picture described in the target video picture carry out face with
Track obtains the second target facial image subset.
Optionally, the above method further include:
Determine the number of the facial image group;
Export the number of the facial image group and the estimation age of each facial image group.
Optionally, above-mentioned quality analysis includes: the quality analysis based on facial angle, the quality based on human face expression point
Analysis, the quality analysis based on face clarity.
According to the second aspect of an embodiment of the present disclosure, a kind of age identification device is provided, comprising:
Target face image set detection module is configured as executing the facial image in detection target video, obtains target
Face image set;
It identifies quality analysis module, is configured as executing each facial image progress year for concentrating the target facial image
Age identification and quality analysis, obtain estimation age and the quality score of each facial image;
Age estimation module, is configured as executing and determines each face according to the estimation age of each facial image and quality score
The estimation age of the affiliated personnel of image.
Optionally, above-mentioned identification quality analysis module includes:
Facial image group generates submodule, is configured as executing to target face image set progress face cluster, obtain
To at least one facial image group;
It identifies quality analysis submodule, is configured as executing for each facial image group, respectively to the facial image group
In each facial image carry out age identification and quality analysis, obtain estimation age and the quality score of each facial image.
Optionally, above-mentioned age estimation module includes:
Age estimates submodule, is configured as executing for each facial image group, according in the facial image group each one
The estimation age of face image and quality score determine the estimation age of the facial image group, wherein each facial image
Each facial image in group belongs to same personnel.
Optionally, the above-mentioned age estimates submodule, comprising:
Weight age computing unit, be configured as executing for each facial image group, by the facial image group each one
The estimation age of face image is multiplied with quality score, obtains the weighting age of each facial image;
Age estimation unit is configured as executing for each facial image group, by face figure each in the facial image group
The weighting age of picture is added, and obtains the estimation age of the facial image group.
Optionally, above-mentioned target face image set includes first object facial image subset, the target face image set
Detection module, comprising:
First refers to facial image detection sub-module, is configured as executing the first frame picture progress to the target video
Face datection obtains first with reference to facial image;
First face tracking submodule is configured as executing according to described first regarding the target with reference to facial image
Other pictures of frequency carry out face tracking, obtain first object facial image subset.
Optionally, above-mentioned target face image set further includes the second target facial image subset, described device further include:
Second refer to facial image detection sub-module, if be configured as execute to the frame picture in other described pictures into
The failure of row face tracking, then carry out Face datection to the picture, and second is obtained when detecting face with reference to facial image;
Second face tracking submodule is configured as executing according to described second regarding the target with reference to facial image
Picture after picture described in frequency carries out face tracking and obtains the second target facial image subset.
Optionally, above-mentioned apparatus further include:
Face number determining module is configured as executing the number for determining the facial image group;
Age output module is configured as executing the estimation of the number and each facial image group that export the facial image group
Age.
Optionally, above-mentioned quality analysis includes: the quality analysis based on facial angle, the quality based on human face expression point
Analysis, the quality analysis based on face clarity.
According to the third aspect of an embodiment of the present disclosure, a kind of electronic equipment is provided, comprising:
Processor, memory and it is stored in the computer journey that can be run on the memory and on the processor
Sequence, the processor realize one or more above-mentioned age recognition methods when executing described program.
According to a fourth aspect of embodiments of the present disclosure, a kind of non-transitorycomputer readable storage medium is provided, when described
When instruction in storage medium is executed by the processor of mobile terminal so that mobile terminal be able to carry out it is one or more above-mentioned
Age recognition methods.
According to a fifth aspect of the embodiments of the present disclosure, a kind of application program/computer program product is provided, comprising:
The facial image in target video is detected, target face image set is obtained;
Age identification and quality analysis are carried out to each facial image that the target facial image is concentrated, obtain each face figure
The estimation age of picture and quality score;
The estimation age of the affiliated personnel of each facial image is determined according to the estimation age of each facial image and quality score.
The technical scheme provided by this disclosed embodiment can include the following benefits: solve age recognition accuracy
Relatively low problem can consider the influence of picture quality when carrying out age identification, and the accuracy of age identification can be improved.
It should be understood that above general description and following detailed description be only it is exemplary and explanatory, not
The disclosure can be limited.
Detailed description of the invention
The drawings herein are incorporated into the specification and forms part of this specification, and shows and meets implementation of the invention
Example, and be used to explain the principle of the present invention together with specification.
Fig. 1 is a kind of flow chart of age recognition methods shown according to an exemplary embodiment;
Fig. 2 is the flow chart of another age recognition methods shown according to an exemplary embodiment.
Fig. 3 is a kind of block diagram of age identification device shown according to an exemplary embodiment;
Fig. 4 is the block diagram of another age identification device shown according to an exemplary embodiment;
Fig. 5 is the block diagram of a kind of electronic equipment shown according to an exemplary embodiment.
Fig. 6 is a kind of block diagram of device for age identification shown according to an exemplary embodiment.
Specific embodiment
Example embodiments are described in detail here, and the example is illustrated in the accompanying drawings.Following description is related to
When attached drawing, unless otherwise indicated, the same numbers in different drawings indicate the same or similar elements.Following exemplary embodiment
Described in embodiment do not represent all embodiments consistented with the present invention.On the contrary, they be only with it is such as appended
The example of device and method being described in detail in claims, some aspects of the invention are consistent.
Fig. 1 is a kind of flow chart of age recognition methods shown according to an exemplary embodiment, as shown in Figure 1, the age
Recognition methods is for including the following steps in terminal.
In a step 101, the facial image in target video is detected, target face image set is obtained.
Wherein, target video can be shot immediately, other clients are sent etc..
Specifically, firstly, detecting the face image set in the first frame image in target video;Then, further according to the people
Face image collection obtains other facial images of different conditions different angle from subsequent image;Finally, from all of target video
All people's face image is got in image, as target face image set.
In a step 102, age identification and quality analysis are carried out to each facial image that the target facial image is concentrated,
Obtain estimation age and the quality score of each facial image.
Wherein, age identification can obtain the age by each characteristic information comprehensive analysis to facial image.For example, can be with
The information such as percutaneous smoothness and muscle thickness determine the age.The algorithm that the embodiment of the present invention identifies the age is not
It limits.
Optionally, in another embodiment of the invention, quality analysis includes but is not limited to: the matter based on facial angle
Measure analysis, the quality analysis based on human face expression, the quality analysis based on face clarity.
It is appreciated that the better picture quality of clarity is better, the poorer picture quality of clarity is poorer, it may be assumed that clarity
With picture quality at positive relationship.Specifically, it is first determined the calculation formula of clarity;Then according to each pixel of facial image
Value calculates clarity;Finally, clarity is normalized to obtain the quality score of facial image.
Facial angle can for shooting figure as when, angle between face direction and the shooting direction of camera.Work as face
When direction is vertical with shooting direction, picture quality is poorer;When face direction and shooting direction are at 180 degree angle, face is being shot just
Face image, picture quality are better.It is appreciated that facial angle make shooting obtain face front portion it is more when, picture quality
Better;Facial angle make shooting obtain face front portion it is fewer when, picture quality is poorer.Specifically, face is defined first
The value rule of angle, then identifies the facial angle in facial image, finally is normalized to obtain by facial angle
The quality score of facial image.
It is appreciated that picture quality is poorer when human face expression makes changes in faces larger;When human face expression makes face
When changing smaller or constant, picture quality is better.For example, can be clearly seen that face are believed when human face expression is to smile
Breath, picture quality are preferable;When human face expression is very ferocious, it cannot be clear that face information, picture quality are poor.
It is appreciated that the embodiment of the present invention is without restriction to the method for quality analysis.In practical applications, it can also adopt
Quality score is determined with the method for other quality analyses, and in conjunction with a variety of analysis algorithms, to improve the accurate of quality analysis
Degree.
In step 103, the affiliated personnel of each facial image are determined according to the estimation age of each facial image and quality score
The estimation age.
Specifically, quality score is lower, indicates that quality of human face image is poorer, so that the estimation age of the facial image is to people
The estimation age effects of face image group are smaller;Quality score is higher, indicates that quality of human face image is better, thus the facial image
Estimate that the age is larger to the estimation age effects of facial image group.Facial image so as to improve high-quality determines to the age
It is qualitative, ropy facial image is reduced to decisive, the final realization accurate estimation age at age.
In embodiments of the present invention, the face image set in target video is detected, target face image set is obtained;To described
Each facial image that target facial image is concentrated carries out age identification and quality analysis, obtain each facial image the estimation age and
Quality score;The estimation age of the affiliated personnel of each facial image is determined according to the estimation age of each facial image and quality score.
It solves the problems, such as that age recognition accuracy is relatively low, the influence of picture quality can be considered when carrying out age identification, it can be with
Improve the accuracy of age identification.
Fig. 2 is the flow chart of another age recognition methods shown according to an exemplary embodiment, as shown in Fig. 2, year
Age recognition methods is for including the following steps in terminal.
In step 201, Face datection is carried out to the first frame picture of the target video, obtains first and refers to face figure
Picture.
Wherein, first can be one or more facial images with reference to facial image.
In practical applications, cascade deep convolutional network model, Faster-RCNN (Faster Region can be passed through
Convolutional Neural Networks, faster region convolutional neural networks) model Face datection, be based on SSD
The Face datections skills such as the Face datection of (Single Shot Multi Box Detector, single-lens more box detectors) model
Art.
It is appreciated that the embodiment of the present invention is without restriction to human face detection tech.
In step 202, facial image is referred to according to described first, face is carried out to other pictures of the target video
Tracking, obtains first object facial image subset.
Specifically, obtain same people's corresponding with the first reference facial image from other pictures using Face tracking algorithm
Facial image.
Face tracking algorithm includes but is not limited to: tracking based on model following, based on motion information, is special based on face part
Sign tracking is tracked based on neural network.
It is exactly the priori knowledge for obtaining target based on model following, establishes low price parameter model, to each frame figure of input
As carrying out Model Matching by sliding window, recognition of face and tracking are realized.Common trace model has: complexion model, ellipse
Model, texture model and eyes template etc..Wherein, the tracking based on complexion model is exactly to utilize appropriate colour of skin system,
Using the colour of skin as the key message for realizing face tracking.Due to the colour of skin have to amplification, reduce and to micro-strain it is insensitive
The advantages of, and influence of the face with respect to the variation of camera lens to Skin Color Information itself is little, so that this method is easy in former frame
The human face region that latter frame image is traced on the basis of image analysis results has the characteristics that speed is fast, posture invariance.Mesh
Preceding face tracking technology mostly uses the method based on complexion model.
The continuity rule that target moves between image successive frame is mainly made full use of based on motion information tracking, is carried out
The prediction of human face region is to achieve the purpose that quickly to track.Generally use the methods of motion segmentation, light stream, stereoscopic vision.It utilizes
Spatio-temporal gradient, Kalman filter are tracked.Light stream is that the pixel movement that space motion object is observed on face generates
Instantaneous velocity field contains the important information of object 3D surface texture and dynamic behaviour.Under normal circumstances, light stream is transported by camera
Dynamic, target movement, or both in scene movement generates.Optical flow analysis is often used in target estimation.When having in scene
When independent moving target, the number of moving target, movement velocity, target range and target can be determined by optical flow analysis
Surface texture.Optical flow analysis can be divided into continuous optical flow method and characteristic light stream method, and characteristic light stream method is acquired by characteristic matching
Light stream at characteristic point.
Organ tracking is carried out according to different human face characteristic informations based on the tracking of face local feature.Such methods warp
Often tracking and positioning is carried out using organ characteristics' information such as eyes, mouth and nose.Traditional Facial features tracking method is usually
Identification point is drawn in human face to track.
Recognition of face and tracking are carried out by neural network model based on neural network tracking.One of method mentions first
50 pivots for taking face, are then mapped it in 5 dimension spaces with auto-correlation neural network;Finally with a common multilayer
Perceptron is differentiated.Another carries out face tracking by hybrid neural network, wherein non-supervisory neural network is used for
Feature extraction, and neural network is supervised for classifying.
In step 203, if face tracking failure is carried out to the frame picture in other described pictures, to the picture
Face datection is carried out, second is obtained when detecting face with reference to facial image.
Specifically, when that can not be traced into a frame picture with the first facial image corresponding with reference to facial image, face
Tracking failure shows that a face tracking is completed.At this point, re-starting Face datection from the picture, new reference man is obtained
Face image, it may be assumed that second refers to facial image.
In step 204, facial image is referred to according to described second, to the figure after picture described in the target video
Piece carries out face tracking and obtains the second target facial image subset.
Face tracking is carried out from other pictures after the picture of face tracking failure, obtains new face image set
It closes, i.e. the second target facial image subset.And so on, until last frame picture.To first object facial image subset
And/or at least one second target facial image subset forms target facial image subset, it may be assumed that the institute for including in the target video
There is facial image.
In step 205, face cluster is carried out to the target face image set, obtains at least one facial image group.
Wherein, the facial image of the different angle of the same person, different time can be divided to a face figure by cluster
As group.
It in practical applications, can be using K-means cluster, hierarchical clustering scheduling algorithm.The embodiment of the present invention is to cluster
Algorithm is without restriction.
Wherein, center of the K-means algorithm any K object of selection random first as initial clustering;Then every
Remaining each object is concentrated to data in secondary iteration, according to it at a distance from the center of each initial clustering, by each object
It reclassifies to nearest cluster.After having sorted out all objects, an iteration is completed, and generates new cluster centre.In addition, such as
For fruit before and after an iteration, the value of K illustrates that algorithm has been restrained there is no variation.K-means cluster is used as phase using distance
Like the evaluation index of property, it may be assumed that the distance of two images is small, and similarity is higher;Conversely, distance is bigger, similarity is lower.
Hierarchical clustering is sorted by the similitude between similarity formula calculate node by similarity from high to low,
Gradually reconnect a node.The advantages of algorithm is can to stop dividing at any time.
In step 206, for each facial image group, is respectively carried out to facial image each in the facial image group age
Identification and quality analysis, obtain estimation age and the quality score of each facial image.
In embodiments of the present invention, the corresponding personnel of a facial image group, thus when there are multiple in target video
When personnel, then multiple facial image groups are obtained.Certainly, when there is no facial image groups when personnel, obtained in target video
Number is 0.
Age identification and quality analysis are referred to the detailed description of step 102, are not repeating herein.
In step 207, for each facial image group, according to the estimation of each facial image in the facial image group
Age and quality score determine the estimation age of the facial image group, wherein each facial image in each facial image group
Belong to same personnel.
Specifically, quality score is lower, indicates that quality of human face image is poorer, so that the estimation age of the facial image is to people
The estimation age effects of face image group are smaller;Quality score is higher, indicates that quality of human face image is better, thus the facial image
Estimate that the age is larger to the estimation age effects of facial image group.Facial image so as to improve high-quality determines to the age
It is qualitative, ropy facial image is reduced to decisive, the final realization accurate estimation age at age.
Optionally, in another embodiment of the invention, above-mentioned step by step rapid 207 include sub-step 2071 to 2072:
In sub-step 2071, for each facial image group, it will estimate described in each facial image in the facial image group
The meter age is multiplied with quality score, obtains the weighting age of each facial image.
Specifically, for j-th of facial image group, the weighting age AQ of i-th of facial imagej,iCalculation formula it is as follows:
AQj,i=Agej,i·Qj,i (1)
Wherein, Agej,iFor the estimation age of i-th of facial image in j-th of facial image group, Qj,iFor j-th of face figure
As the quality score of i-th of facial image in group, i and j are the positive integer greater than 0.
In step 2072, for each facial image group, by the weighting age of each facial image in the facial image group
It is added, obtains the estimation age of the facial image group.
Specifically, for j-th of facial image group, estimate age AQjCalculation formula it is as follows:
Wherein, IjFor the corresponding facial image number of j-th of facial image group.
In a step 208, the number of the facial image group is determined.
It is appreciated that the number of facial image group represents the number of people in target video.
In step 209, the number of the facial image group and the estimation age of each facial image group are exported.
In embodiments of the present invention, the number of people and each one can be exported in target video at corresponding estimation age.
It is appreciated that in practical applications, one of facial image conduct can also be chosen from facial image group should
The portrait of people exports, and indicates the estimation age of the people.
In embodiments of the present invention, the facial image in target video is detected, target face image set is obtained;To the mesh
It marks each facial image that facial image is concentrated and carries out age identification and quality analysis, obtain estimation age and the matter of each facial image
It measures point;The estimation age of the affiliated personnel of each facial image is determined according to the estimation age of each facial image and quality score.Solution
It has determined the relatively low problem of age recognition accuracy, the influence of picture quality, Ke Yiti can be considered when carrying out age identification
The accuracy of high age identification.
Fig. 3 is a kind of block diagram of age identification device shown according to an exemplary embodiment.Referring to Fig. 3, the device packet
Include target face image set detection module 301, identification quality analysis module 302 and age estimation module 303.
The target face image set detection module 301 is configured as executing the facial image in detection target video, obtain
Target face image set.
The identification quality analysis module 302 is configured as executing each facial image for concentrating the target facial image
Age identification and quality analysis are carried out, estimation age and the quality score of each facial image are obtained.
The age estimation module 303 is configured as executing determining according to the estimation age of each facial image and quality score
The estimation age of each affiliated personnel of facial image.
In embodiments of the present invention, the facial image in target video is detected, target face image set is obtained;To the mesh
It marks each facial image that facial image is concentrated and carries out age identification and quality analysis, obtain estimation age and the matter of each facial image
It measures point;The estimation age of the affiliated personnel of each facial image is determined according to the estimation age of each facial image and quality score.Solution
It has determined the relatively low problem of age recognition accuracy, the influence of picture quality, Ke Yiti can be considered when carrying out age identification
The accuracy of high age identification.
About the device in above-described embodiment, wherein modules execute the concrete mode of operation in related this method
Embodiment in be described in detail, no detailed explanation will be given here.
Fig. 4 is the block diagram of another age identification device shown according to an exemplary embodiment.Referring to Fig. 4, the device
It is true including target face image set detection module 401, identification quality analysis module 402, age estimation module 403, face number
Cover half block 404 and age output module 405.
The target face image set detection module 401 is configured as executing the facial image in detection target video, obtain
Target face image set.Optionally, in embodiments of the present invention, target face image set detection module 401, comprising: first
Facial image detection sub-module is referred to reference to facial image detection sub-module 4011, the first face tracking submodule 4012, second
4013 and the second face tracking submodule 4014.
The first reference facial image detection sub-module 4011, is configured as executing the first frame figure to the target video
Piece carries out Face datection, obtains first with reference to facial image.
The first face tracking submodule 4012 is configured as executing according to described first with reference to facial image, to described
Other pictures of target video carry out face tracking, obtain first object facial image subset.
The second reference facial image detection sub-module 4013, if being configured as executing to the frame in other described pictures
Picture carries out face tracking failure, then carries out Face datection to the picture, and second is obtained when detecting face with reference to face
Image.
The second face tracking submodule 4014 is configured as executing according to described second with reference to facial image, to described
Picture after picture described in target video carries out face tracking and obtains the second target facial image subset.
The identification quality analysis module 402 is configured as executing each facial image for concentrating the target facial image
Age identification and quality analysis are carried out, estimation age and the quality score of each facial image are obtained.Optionally, implement in the present invention
In example, which includes: that facial image group generates submodule 4021 and identification quality analysis submodule
4022。
The facial image group generates submodule 4021, is configured as executing poly- to target face image set progress face
Class obtains at least one facial image group.
The identification quality analysis submodule 4022 is configured as executing for each facial image group, respectively to the face
Each facial image carries out age identification and quality analysis in image group, obtains estimation age and the quality score of each facial image.
The age estimation module 403 is configured as executing determining according to the estimation age of each facial image and quality score
The estimation age of each affiliated personnel of facial image.Optionally, in embodiments of the present invention, which includes: year
Age estimates submodule 4031.
The age estimates submodule 4031, is configured as executing for each facial image group, according to the facial image group
In each facial image the estimation age and quality score determine estimation age of the facial image group, wherein everyone
Each facial image in face image group belongs to same personnel.
The face number determining module 404 is configured as executing the number for determining the facial image group.
The age output module 405 is configured as executing the number and each facial image group for exporting the facial image group
The estimation age.
Optionally, in another embodiment of the invention, which includes: to calculate list at the weighting age
Member and age estimation unit.
The weighting age computing unit is configured as executing for each facial image group, will be each in the facial image group
The estimation age of facial image is multiplied with quality score, obtains the weighting age of each facial image.
The age estimation unit is configured as executing for each facial image group, by each face in the facial image group
The weighting age of image is added, and obtains the estimation age of the facial image group.
Optionally, in another embodiment of the invention, above-mentioned quality analysis includes: the quality based on facial angle point
Analysis, the quality analysis based on human face expression, the quality analysis based on face clarity.
In embodiments of the present invention, the facial image in target video is detected, target face image set is obtained;To the mesh
It marks each facial image that facial image is concentrated and carries out age identification and quality analysis, obtain estimation age and the matter of each facial image
It measures point;The estimation age of the affiliated personnel of each facial image is determined according to the estimation age of each facial image and quality score.Solution
It has determined the relatively low problem of age recognition accuracy, the influence of picture quality, Ke Yiti can be considered when carrying out age identification
The accuracy of high age identification.
About the device in above-described embodiment, wherein modules execute the concrete mode of operation in related this method
Embodiment in be described in detail, no detailed explanation will be given here.
Fig. 5 is a kind of block diagram for electronic equipment 500 shown according to an exemplary embodiment.For example, electronic equipment
500 can be mobile phone, computer, digital broadcasting terminal, messaging device, game console, tablet device, and medical treatment is set
It is standby, body-building equipment, personal digital assistant etc..
Referring to Fig. 5, electronic equipment 500 may include following one or more components: processing component 502, memory 504,
Electric power assembly 506, multimedia component 508, audio component 510, the interface 512 of input/output (I/O), sensor module 514,
And communication component 516.
The integrated operation of the usual controlling electronic devices 500 of processing component 502, such as with display, call, data are logical
Letter, camera operation and record operate associated operation.Processing component 502 may include one or more processors 520 to hold
Row instruction, to complete all or part of the steps of above-mentioned age recognition methods.In addition, processing component 502 may include one
Or multiple modules, convenient for the interaction between processing component 502 and other assemblies.For example, processing component 502 may include multimedia
Module, to facilitate the interaction between multimedia component 508 and processing component 502.
Memory 504 is configured as storing various types of data to support the operation in equipment 500.These data are shown
Example includes the instruction of any application or method for operating on electronic equipment 500, contact data, telephone directory number
According to, message, picture, video etc..Memory 504 can by any kind of volatibility or non-volatile memory device or they
Combination realize, such as static random access memory (SRAM), electrically erasable programmable read-only memory (EEPROM) is erasable
Programmable read only memory (EPROM), programmable read only memory (PROM), read-only memory (ROM), magnetic memory, quick flashing
Memory, disk or CD.
Power supply module 506 provides electric power for the various assemblies of electronic equipment 500.Power supply module 506 may include power supply pipe
Reason system, one or more power supplys and other with for electronic equipment 500 generate, manage, and distribute the associated component of electric power.
Multimedia component 508 includes the screen of one output interface of offer between the electronic equipment 500 and user.
In some embodiments, screen may include liquid crystal display (LCD) and touch panel (TP).If screen includes touch surface
Plate, screen may be implemented as touch screen, to receive input signal from the user.Touch panel includes one or more touches
Sensor is to sense the gesture on touch, slide, and touch panel.The touch sensor can not only sense touch or sliding
The boundary of movement, but also detect duration and pressure associated with the touch or slide operation.In some embodiments,
Multimedia component 508 includes a front camera and/or rear camera.When equipment 500 is in operation mode, as shot mould
When formula or video mode, front camera and/or rear camera can receive external multi-medium data.Each preposition camera shooting
Head and rear camera can be a fixed optical lens system or have focusing and optical zoom capabilities.
Audio component 510 is configured as output and/or input audio signal.For example, audio component 510 includes a Mike
Wind (MIC), when electronic equipment 500 is in operation mode, when such as call mode, recording mode, and voice recognition mode, microphone
It is configured as receiving external audio signal.The received audio signal can be further stored in memory 504 or via logical
Believe that component 516 is sent.In some embodiments, audio component 510 further includes a loudspeaker, is used for output audio signal.
I/O interface 512 provides interface between processing component 502 and peripheral interface module, and above-mentioned peripheral interface module can
To be keyboard, click wheel, button etc..These buttons may include, but are not limited to: home button, volume button, start button and lock
Determine button.
Sensor module 514 includes one or more sensors, for providing the state of various aspects for electronic equipment 500
Assessment.For example, sensor module 514 can detecte the state that opens/closes of equipment 500, the relative positioning of component, such as institute
The display and keypad that component is electronic equipment 500 are stated, sensor module 514 can also detect electronic equipment 500 or electronics
The position change of 500 1 components of equipment, the existence or non-existence that user contacts with electronic equipment 500,500 orientation of electronic equipment
Or the temperature change of acceleration/deceleration and electronic equipment 500.Sensor module 514 may include proximity sensor, be configured to
It detects the presence of nearby objects without any physical contact.Sensor module 514 can also include optical sensor, such as
CMOS or ccd image sensor, for being used in imaging applications.In some embodiments, which can be with
Including acceleration transducer, gyro sensor, Magnetic Sensor, pressure sensor or temperature sensor.
Communication component 516 is configured to facilitate the communication of wired or wireless way between electronic equipment 500 and other equipment.
Electronic equipment 500 can access the wireless network based on communication standard, such as WiFi, carrier network (such as 2G, 3G, 4G or 5G),
Or their combination.In one exemplary embodiment, communication component 516 receives via broadcast channel and comes from external broadcasting management
The broadcast singal or broadcast related information of system.In one exemplary embodiment, the communication component 516 further includes that near field is logical
(NFC) module is believed, to promote short range communication.For example, radio frequency identification (RFID) technology, infrared data association can be based in NFC module
Meeting (IrDA) technology, ultra wide band (UWB) technology, bluetooth (BT) technology and other technologies are realized.
In the exemplary embodiment, electronic equipment 500 can be by one or more application specific integrated circuit (ASIC), number
Word signal processor (DSP), digital signal processing appts (DSPD), programmable logic device (PLD), field programmable gate array
(FPGA), controller, microcontroller, microprocessor or other electronic components are realized, for executing the above method.
In the exemplary embodiment, a kind of non-transitorycomputer readable storage medium including instruction, example are additionally provided
It such as include the memory 504 of instruction, above-metioned instruction can be executed by the processor 520 of electronic equipment 500 to complete the above method.Example
Such as, the non-transitorycomputer readable storage medium can be ROM, random access memory (RAM), CD-ROM, tape, soft
Disk and optical data storage devices etc..
Fig. 6 is a kind of block diagram of device 600 for age identification shown according to an exemplary embodiment.For example, dress
Setting 600 may be provided as a server.Referring to Fig. 6, device 600 includes processing component 622, further comprises one or more
A processor, and the memory resource as representated by memory 632, can be by the finger of the execution of processing component 622 for storing
It enables, such as application program.The application program stored in memory 632 may include it is one or more each correspond to
The module of one group of instruction.In addition, processing component 622 is configured as executing instruction, to execute the above method.
Device 600 can also include the power management that a power supply module 626 is configured as executive device 600, and one has
Line or radio network interface 650 are configured as device 600 being connected to network and input and output (I/O) interface 658.Dress
Setting 600 can operate based on the operating system for being stored in memory 632, such as Windows ServerTM, Mac OS XTM,
UnixTM, LinuxTM, FreeBSDTM or similar.
The embodiment of the present disclosure additionally provides a kind of non-transitorycomputer readable storage medium, when in the storage medium
When instruction is executed by the processor of mobile terminal, so that mobile terminal is able to carry out one or more above-mentioned age identification sides
Method.
The embodiment of the present disclosure also provides a kind of application program/computer program product, comprising:
The facial image in target video is detected, target face image set is obtained;
Age identification and quality analysis are carried out to each facial image that the target facial image is concentrated, obtain each face figure
The estimation age of picture and quality score;
The estimation age of the affiliated personnel of each facial image is determined according to the estimation age of each facial image and quality score.
Those skilled in the art after considering the specification and implementing the invention disclosed here, will readily occur to of the invention its
Its embodiment.This application is intended to cover any variations, uses, or adaptations of the invention, these modifications, purposes or
Person's adaptive change follows general principle of the invention and including the undocumented common knowledge in the art of the disclosure
Or conventional techniques.The description and examples are only to be considered as illustrative, and true scope and spirit of the invention are by following
Claim is pointed out.
It should be understood that the present invention is not limited to the precise structure already described above and shown in the accompanying drawings, and
And various modifications and changes may be made without departing from the scope thereof.The scope of the present invention is limited only by the attached claims.
Claims (10)
1. a kind of age recognition methods, which is characterized in that the described method includes:
The facial image in target video is detected, target face image set is obtained;
Age identification and quality analysis are carried out to each facial image that the target facial image is concentrated, obtain each facial image
Estimate age and quality score;
The estimation age of the affiliated personnel of each facial image is determined according to the estimation age of each facial image and quality score.
2. the method according to claim 1, wherein each face figure concentrated to the target facial image
As carrying out age identification and quality analysis, estimation age and the quality score of each facial image are obtained, comprising:
Face cluster is carried out to the target face image set, obtains at least one facial image group;
For each facial image group, age identification and quality analysis are carried out to facial image each in the facial image group respectively,
Obtain estimation age and the quality score of each facial image.
3. according to the method described in claim 2, it is characterized in that, described obtain according to the estimation age of each facial image and quality
Divide the estimation age for determining each affiliated personnel of facial image, comprising:
It is true according to the estimation age of each facial image in the facial image group and quality score for each facial image group
The estimation age of the fixed facial image group, wherein each facial image in each facial image group belongs to same personnel.
4. according to the method described in claim 3, it is characterized in that, described for each facial image group, according to the face figure
As the estimation age of facial image each in group and quality score determine the estimation age of the facial image group, comprising:
For each facial image group, by the estimation age of each facial image in the facial image group and quality score phase
Multiply, obtains the weighting age of each facial image;
For each facial image group, the weighting age of each facial image in the facial image group is added, the face is obtained
The estimation age of image group.
5. the method according to claim 1, wherein the target face image set includes first object face figure
As subset, the facial image detected in target video obtains target face image set, comprising:
Face datection is carried out to the first frame picture of the target video, obtains first with reference to facial image;
Facial image is referred to according to described first, face tracking is carried out to remaining picture of the target video, obtains the first mesh
Mark facial image subset.
6. according to the method described in claim 5, it is characterized in that, the target face image set further includes the second target face
Image subset, the method also includes:
If carrying out face tracking failure to the frame picture in remaining picture of the target video, people is carried out to the picture
Face detection obtains second when detecting face with reference to facial image;
Facial image is referred to according to described second, face tracking is carried out to the picture after picture described in the target video and is obtained
To the second target facial image subset.
7. the method according to claim 1, wherein the method also includes:
Determine the number of the facial image group;
Export the number of the facial image group and the estimation age of each facial image group.
8. a kind of age identification device characterized by comprising
Target face image set detection module is configured as executing the facial image in detection target video, obtains target face
Image set;
It identifies quality analysis module, is configured as executing each facial image progress age knowledge for concentrating the target facial image
Other and quality analysis, obtains estimation age and the quality score of each facial image;
Age estimation module, is configured as executing and determines each facial image according to the estimation age of each facial image and quality score
The estimation age of affiliated personnel.
9. a kind of electronic equipment characterized by comprising
Processor, memory and it is stored in the computer program that can be run on the memory and on the processor, institute
State the age recognition methods realized as described in one or more in claim 1 to 7 when processor executes described program.
10. a kind of non-transitorycomputer readable storage medium, which is characterized in that when the instruction in the storage medium is by electronics
When the processor of equipment executes, so that electronic equipment is able to carry out the year as described in one or more in claim to a method 1 to 7
Age recognition methods.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810747040.0A CN109190449A (en) | 2018-07-09 | 2018-07-09 | Age recognition methods, device, electronic equipment and storage medium |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810747040.0A CN109190449A (en) | 2018-07-09 | 2018-07-09 | Age recognition methods, device, electronic equipment and storage medium |
Publications (1)
Publication Number | Publication Date |
---|---|
CN109190449A true CN109190449A (en) | 2019-01-11 |
Family
ID=64936387
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810747040.0A Pending CN109190449A (en) | 2018-07-09 | 2018-07-09 | Age recognition methods, device, electronic equipment and storage medium |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN109190449A (en) |
Cited By (16)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110020638A (en) * | 2019-04-17 | 2019-07-16 | 唐晓颖 | Facial expression recognizing method, device, equipment and medium |
CN110163171A (en) * | 2019-05-27 | 2019-08-23 | 北京字节跳动网络技术有限公司 | The method and apparatus of face character for identification |
CN110163170A (en) * | 2019-05-27 | 2019-08-23 | 北京字节跳动网络技术有限公司 | The method and apparatus at age for identification |
CN110175565A (en) * | 2019-05-27 | 2019-08-27 | 北京字节跳动网络技术有限公司 | The method and apparatus of personage's emotion for identification |
CN110188660A (en) * | 2019-05-27 | 2019-08-30 | 北京字节跳动网络技术有限公司 | The method and apparatus at age for identification |
CN110334663A (en) * | 2019-07-09 | 2019-10-15 | 北京字节跳动网络技术有限公司 | Age recognition methods and device, storage medium and terminal based on image |
CN110399839A (en) * | 2019-07-26 | 2019-11-01 | 北京达佳互联信息技术有限公司 | Face identification method, device, equipment and storage medium |
CN110838102A (en) * | 2019-10-24 | 2020-02-25 | 开望(杭州)科技有限公司 | Intelligent image uploading method |
CN111407245A (en) * | 2020-03-19 | 2020-07-14 | 南京昊眼晶睛智能科技有限公司 | Non-contact heart rate and body temperature measuring method based on camera |
CN111459587A (en) * | 2020-03-27 | 2020-07-28 | 北京三快在线科技有限公司 | Information display method, device, equipment and storage medium |
CN111666908A (en) * | 2020-06-09 | 2020-09-15 | 广州市百果园信息技术有限公司 | Interest portrait generation method, device and equipment for video user and storage medium |
CN112132068A (en) * | 2020-09-27 | 2020-12-25 | 深圳市梦网视讯有限公司 | Age analysis method, system and equipment based on video dynamic information |
CN112188091A (en) * | 2020-09-24 | 2021-01-05 | 北京达佳互联信息技术有限公司 | Face information identification method and device, electronic equipment and storage medium |
WO2021012383A1 (en) * | 2019-07-24 | 2021-01-28 | 初速度(苏州)科技有限公司 | Age prediction method and device for infrared image |
CN112766238A (en) * | 2021-03-15 | 2021-05-07 | 电子科技大学中山学院 | Age prediction method and device |
CN113837043A (en) * | 2021-09-14 | 2021-12-24 | 汇纳科技股份有限公司 | Age estimation method, device and system applied to face image and storage medium |
Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104036236A (en) * | 2014-05-27 | 2014-09-10 | 厦门瑞为信息技术有限公司 | Human face gender recognition method based on multi-parameter index weighting |
KR20160110741A (en) * | 2015-03-11 | 2016-09-22 | 동국대학교 산학협력단 | Device and method for human age estimation |
CN106295499A (en) * | 2016-07-21 | 2017-01-04 | 北京小米移动软件有限公司 | Age estimation method and device |
CN107133576A (en) * | 2017-04-17 | 2017-09-05 | 北京小米移动软件有限公司 | Age of user recognition methods and device |
CN108229322A (en) * | 2017-11-30 | 2018-06-29 | 北京市商汤科技开发有限公司 | Face identification method, device, electronic equipment and storage medium based on video |
-
2018
- 2018-07-09 CN CN201810747040.0A patent/CN109190449A/en active Pending
Patent Citations (5)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN104036236A (en) * | 2014-05-27 | 2014-09-10 | 厦门瑞为信息技术有限公司 | Human face gender recognition method based on multi-parameter index weighting |
KR20160110741A (en) * | 2015-03-11 | 2016-09-22 | 동국대학교 산학협력단 | Device and method for human age estimation |
CN106295499A (en) * | 2016-07-21 | 2017-01-04 | 北京小米移动软件有限公司 | Age estimation method and device |
CN107133576A (en) * | 2017-04-17 | 2017-09-05 | 北京小米移动软件有限公司 | Age of user recognition methods and device |
CN108229322A (en) * | 2017-11-30 | 2018-06-29 | 北京市商汤科技开发有限公司 | Face identification method, device, electronic equipment and storage medium based on video |
Cited By (21)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110020638A (en) * | 2019-04-17 | 2019-07-16 | 唐晓颖 | Facial expression recognizing method, device, equipment and medium |
CN110163171A (en) * | 2019-05-27 | 2019-08-23 | 北京字节跳动网络技术有限公司 | The method and apparatus of face character for identification |
CN110163170A (en) * | 2019-05-27 | 2019-08-23 | 北京字节跳动网络技术有限公司 | The method and apparatus at age for identification |
CN110175565A (en) * | 2019-05-27 | 2019-08-27 | 北京字节跳动网络技术有限公司 | The method and apparatus of personage's emotion for identification |
CN110188660A (en) * | 2019-05-27 | 2019-08-30 | 北京字节跳动网络技术有限公司 | The method and apparatus at age for identification |
WO2020238321A1 (en) * | 2019-05-27 | 2020-12-03 | 北京字节跳动网络技术有限公司 | Method and device for age identification |
CN110334663A (en) * | 2019-07-09 | 2019-10-15 | 北京字节跳动网络技术有限公司 | Age recognition methods and device, storage medium and terminal based on image |
WO2021012383A1 (en) * | 2019-07-24 | 2021-01-28 | 初速度(苏州)科技有限公司 | Age prediction method and device for infrared image |
CN110399839A (en) * | 2019-07-26 | 2019-11-01 | 北京达佳互联信息技术有限公司 | Face identification method, device, equipment and storage medium |
CN110399839B (en) * | 2019-07-26 | 2021-07-16 | 北京达佳互联信息技术有限公司 | Face recognition method, device, equipment and storage medium |
CN110838102A (en) * | 2019-10-24 | 2020-02-25 | 开望(杭州)科技有限公司 | Intelligent image uploading method |
CN111407245A (en) * | 2020-03-19 | 2020-07-14 | 南京昊眼晶睛智能科技有限公司 | Non-contact heart rate and body temperature measuring method based on camera |
CN111459587A (en) * | 2020-03-27 | 2020-07-28 | 北京三快在线科技有限公司 | Information display method, device, equipment and storage medium |
CN111666908A (en) * | 2020-06-09 | 2020-09-15 | 广州市百果园信息技术有限公司 | Interest portrait generation method, device and equipment for video user and storage medium |
CN112188091A (en) * | 2020-09-24 | 2021-01-05 | 北京达佳互联信息技术有限公司 | Face information identification method and device, electronic equipment and storage medium |
CN112188091B (en) * | 2020-09-24 | 2022-05-06 | 北京达佳互联信息技术有限公司 | Face information identification method and device, electronic equipment and storage medium |
CN112132068A (en) * | 2020-09-27 | 2020-12-25 | 深圳市梦网视讯有限公司 | Age analysis method, system and equipment based on video dynamic information |
CN112766238A (en) * | 2021-03-15 | 2021-05-07 | 电子科技大学中山学院 | Age prediction method and device |
CN112766238B (en) * | 2021-03-15 | 2023-09-26 | 电子科技大学中山学院 | Age prediction method and device |
CN113837043A (en) * | 2021-09-14 | 2021-12-24 | 汇纳科技股份有限公司 | Age estimation method, device and system applied to face image and storage medium |
CN113837043B (en) * | 2021-09-14 | 2023-11-21 | 汇纳科技股份有限公司 | Age estimation method, device and system applied to face image and storage medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN109190449A (en) | Age recognition methods, device, electronic equipment and storage medium | |
CN109063580A (en) | Face identification method, device, electronic equipment and storage medium | |
CN106295566B (en) | Facial expression recognizing method and device | |
TWI766286B (en) | Image processing method and image processing device, electronic device and computer-readable storage medium | |
CN108960209B (en) | Identity recognition method, identity recognition device and computer readable storage medium | |
CN105631403B (en) | Face identification method and device | |
CN110602527B (en) | Video processing method, device and storage medium | |
CN105608425B (en) | The method and device of classification storage is carried out to photo | |
CN104408402B (en) | Face identification method and device | |
CN110544272B (en) | Face tracking method, device, computer equipment and storage medium | |
CN106295511B (en) | Face tracking method and device | |
CN110516745A (en) | Training method, device and the electronic equipment of image recognition model | |
CN105654039B (en) | The method and apparatus of image procossing | |
US20120321193A1 (en) | Method, apparatus, and computer program product for image clustering | |
CN107368810A (en) | Method for detecting human face and device | |
CN107463903B (en) | Face key point positioning method and device | |
Kang et al. | Development of head detection and tracking systems for visual surveillance | |
CN107688781A (en) | Face identification method and device | |
CN111985268A (en) | Method and device for driving animation by human face | |
CN109543066A (en) | Video recommendation method, device and computer readable storage medium | |
CN105528078B (en) | The method and device of controlling electronic devices | |
CN104408404A (en) | Face identification method and apparatus | |
CN109360197A (en) | Processing method, device, electronic equipment and the storage medium of image | |
CN109242045B (en) | Image clustering processing method, device, electronic equipment and storage medium | |
CN109670077A (en) | Video recommendation method, device and computer readable storage medium |
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 |