CN108388889A - Method and apparatus for analyzing facial image - Google Patents
Method and apparatus for analyzing facial image Download PDFInfo
- Publication number
- CN108388889A CN108388889A CN201810247286.1A CN201810247286A CN108388889A CN 108388889 A CN108388889 A CN 108388889A CN 201810247286 A CN201810247286 A CN 201810247286A CN 108388889 A CN108388889 A CN 108388889A
- Authority
- CN
- China
- Prior art keywords
- face
- image
- target object
- face depth
- analysis result
- 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.)
- Granted
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
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F18/00—Pattern recognition
- G06F18/20—Analysing
- G06F18/22—Matching criteria, e.g. proximity measures
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/50—Depth or shape recovery
-
- 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
- G06V40/168—Feature extraction; Face representation
- G06V40/169—Holistic features and representations, i.e. based on the facial image taken as a whole
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
Abstract
The embodiment of the present application discloses the method and apparatus for analyzing facial image.One specific implementation mode of this method includes：Obtain the face flat image and face depth image of target object；According to face flat image, simulation face depth data of the generation target object in preset posture；Simulation face depth data is subjected to the matching analysis with face depth image；According to the matching analysis as a result, generating the human face analysis result of target object.The embodiment can utilize face flat image to generate simulation face depth data, so as to carry out the matching analysis with the face depth image of acquisition, to determine whether the face depth image comes from real human face.This method can be applied to face In vivo detection.
Description
Technical field
The invention relates to field of computer technology, and in particular to the method and apparatus for analyzing facial image.
Background technology
Recognition of face, the facial feature information for being normally based on people carry out a kind of biological identification technology of identification.With
Video camera or camera acquire image or video flowing containing face, and automatic detect and track face in the picture, and then right
The face that detects carries out a series of the relevant technologies analysis of face, usually also referred to as Identification of Images, face recognition.
In actual application, we can not only collect face flat image, can also collect face depth
Image.Meanwhile we generally require to judge collected object whether be really be face, i.e., whether meet stereoscopic face structure (three
Tie up human face structure), to realize that live body (people for having vital sign) detects.
Invention content
The embodiment of the present application proposes the method and apparatus for analyzing facial image.
In a first aspect, the embodiment of the present application provides a kind of method for analyzing facial image, including：Obtain target pair
The face flat image and face depth image of elephant；According to face flat image, simulation of the target object in preset posture is generated
Face depth data；Simulation face depth data is subjected to the matching analysis with face depth image；It is according to the matching analysis as a result, raw
At the human face analysis result of target object.
In some embodiments, according to face flat image, simulation face depth of the generation target object in preset posture
Data, including：Face flat image is inputted to the three-dimensional deformation model built in advance, generates target object the three of preset posture
Tie up faceform；According to three-dimensional face model, face depth data of the extraction target object under preset posture generates simulation people
Face depth data.
In some embodiments, according to the matching analysis as a result, generating the human face analysis of target object as a result, including：It determines
Whether the matching analysis result meets preset condition；In response to determining that the matching analysis result meets preset condition, generate for characterizing
The face of target object meets the analysis result of the stereoscopic face structure under preset posture；In response to determining the matching analysis result not
Meet preset condition, generates the analysis knot for not meeting the stereoscopic face structure under preset posture for characterizing the face of target object
Fruit.
In some embodiments, face depth image is obtained by following steps：Depth image collecting device is received to be adopted
The face depth image of the target object of collection；Or receive the people for the target object that at least two image capture devices acquire simultaneously
Face image；According to the facial image of reception, the face depth image of target object is generated.
In some embodiments, according to the facial image of reception, the face depth image of target object is generated, including：Root
According to the facial image of reception, determination is currently located at whether the face of the target object before image capture device is stereochemical structure；It rings
The face depth image of target object should be generated in determining that the face of target object is stereochemical structure.
In some embodiments, face flat image includes following at least one：Near-infrared image, gray level image and colour
Image.
Second aspect, the embodiment of the present application provide a kind of device for analyzing facial image, including：Acquiring unit,
It is configured to obtain the face flat image and face depth image of target object；Data generating unit is configured to according to people
Face flat image, simulation face depth data of the generation target object in preset posture；Analytic unit is configured to that people will be simulated
Face depth data carries out the matching analysis with face depth image；As a result generation unit is configured to according to the matching analysis as a result, raw
At the human face analysis result of target object.
In some embodiments, data generating unit includes：First generates subelement, is configured to face flat image
Input the three-dimensional deformation model built in advance, three-dimensional face model of the generation target object in preset posture；Second generates son list
Member is configured to according to three-dimensional face model, and face depth data of the extraction target object under preset posture generates simulation people
Face depth data.
In some embodiments, as a result generation unit includes：Determination subelement, be configured to determine the matching analysis the result is that
It is no to meet preset condition；First response subelement is configured to, in response to determining that the matching analysis result meets preset condition, generate
Face for characterizing target object meets the analysis result of the stereoscopic face structure under preset posture；Second response subelement,
It is configured in response to determining that the matching analysis result is unsatisfactory for preset condition, generation is not met for characterizing the face of target object
The analysis result of stereoscopic face structure under preset posture.
In some embodiments, which further includes image generation unit, is configured to：Receive depth image collecting device
The face depth image of the target object acquired；Or receive the target object that at least two image capture devices acquire simultaneously
Facial image；According to the facial image of reception, the face depth image of target object is generated.
In some embodiments, image generation unit is further configured to：According to the facial image of reception, determine current
Whether the face of the target object before image capture device is stereochemical structure；In response to determining that the face of target object is vertical
Body structure generates the face depth image of target object.
In some embodiments, face flat image includes following at least one：Near-infrared image, gray level image and colour
Image.
The third aspect, the embodiment of the present application provide a kind of electronic equipment, including：One or more processors；Storage dress
It sets, for storing one or more programs；When one or more programs are executed by one or more processors so that one or more
A processor realizes the method as described in any embodiment in above-mentioned first aspect.
Fourth aspect, the embodiment of the present application provide a kind of computer-readable medium, are stored thereon with computer program,
In, the method as described in any embodiment in above-mentioned first aspect is realized when which is executed by processor.
Method and apparatus provided by the embodiments of the present application for analyzing facial image, by the face for obtaining target object
Flat image and face depth image, so as to according to the face flat image of acquisition, generate target object identical default
Simulation face depth data under posture；Later, the face depth map of simulation face depth data and acquisition will can be generated
As carrying out the matching analysis；Finally, according to the matching analysis as a result, the human face analysis result of target object can be generated.Utilize people
Face flat image generates simulation face depth data, so as to carry out the matching analysis with the face depth image of acquisition, with true
Whether the fixed face depth image comes from real human face.This method and apparatus can be applied to face In vivo detection.
Description of the drawings
By reading a detailed description of non-restrictive embodiments in the light of the attached drawings below, the application's is other
Feature, objects and advantages will become more apparent upon：
Fig. 1 is that this application can be applied to exemplary system architecture figures therein；
Fig. 2 is the flow chart of the one embodiment for the method for being used to analyze facial image according to the application；
Fig. 3 is the schematic diagram of an application scenarios of the method for being used to analyze facial image according to the application；
Fig. 4 is the structural schematic diagram of the one embodiment for the device for being used to analyze facial image according to the application；
Fig. 5 is adapted for the structural schematic diagram of the computer system of the electronic equipment for realizing the embodiment of the present application.
Specific implementation mode
The application is described in further detail with reference to the accompanying drawings and examples.It is understood that this place is retouched
The specific embodiment stated is used only for explaining related invention, rather than the restriction to the invention.It also should be noted that in order to
Convenient for description, is illustrated only in attached drawing and invent relevant part with related.
It should be noted that in the absence of conflict, the features in the embodiments and the embodiments of the present application can phase
Mutually combination.The application is described in detail below with reference to the accompanying drawings and in conjunction with the embodiments.
Fig. 1 is shown can be using the embodiment of the present application for analyzing the method for facial image or for analyzing face figure
The exemplary system architecture 100 of the device of picture.
As shown in Figure 1, system architecture 100 may include terminal 101, network 102 and server 103.Network 102 to
The medium of communication link is provided between terminal 101 and server 103.Network 102 may include various connection types, such as have
Line, wireless communication link or fiber optic cables etc..
User can be interacted with using terminal 101 by network 102 and server 103, to receive or send message etc..
Various client applications, such as the application of facial image analysis and identification class, the application of shopping class, payment can be installed in terminal 101
Class application, web browser and immediate communication tool etc..
Here terminal 101 can be hardware, can also be software.Can be that there is display when terminal 101 is hardware
The various electronic equipments of screen, including but not limited to smart mobile phone, tablet computer, E-book reader, MP3 player (Moving
Picture Experts Group Audio Layer III, dynamic image expert's compression standard audio level 3), it is on knee just
Take computer and desktop computer etc..When terminal 101 is software, may be mounted in above-mentioned cited electronic equipment.
Multiple softwares or software module (such as providing Distributed Services) may be implemented into it, can also be implemented as single software or
Software module.It is not specifically limited herein.
When terminal 101 is hardware, it is also equipped with image capture device 1011,1012 thereon.Image capture device
1011,1012 different location that may be mounted at terminal 101, the face figure of the different angle for acquiring same target object
Picture.Image capture device 1011,1012 can be the various equipment that can realize Image Acquisition, such as camera, sensor.
Server 103 can be to provide the server of various services, such as the various applications to being shown in terminal 101 provide
The background server of support.Background server can send terminal 101, image capture device 1011,1012 is acquired
The facial image of target object carries out analyzing processing, and can handling result be sent to end (such as the human face analysis result generated)
End 101.In this way, can make terminal 101 that handling result is showed user, or the application in terminal 101 can be made according to place
It manages result and realizes corresponding feature operation.
Here server 103 can be equally hardware, can also be software.When server 103 is hardware, Ke Yishi
The distributed server cluster of ready-made multiple server compositions, can also be implemented as individual server.When server 103 is software
When, multiple softwares or software module (such as providing Distributed Services) may be implemented into, single software can also be implemented as
Or software module.It is not specifically limited herein.
It should be noted that the method for analyzing facial image that the embodiment of the present application is provided is generally by server
103 execute.Correspondingly, it is generally positioned in server 103 for analyzing the device of facial image.
It should be pointed out that image capture device 1011,1012 may be mounted in terminal 101, it can also be independently of end
It holds 101 and is separately provided.And when image capture device is independently arranged, it again may be by network 102 and realized with server 103
Communication.
It should be understood that the number of the terminal, network, server and image capture device in Fig. 1 is only schematical.Root
It factually now needs, can have any number of terminal, network, server and image capture device.
With continued reference to Fig. 2, it illustrates according to one embodiment of the method for analyzing facial image of the application
Flow 200.The method for being used to analyze facial image may comprise steps of：
Step 201, the face flat image and face depth image of target object are obtained.
In the present embodiment, the executive agent (such as server 103 shown in FIG. 1) of the method for analyzing facial image
The face flat image and face depth image of target object can be obtained in several ways.For example, executive agent can be with
By wired connection mode or radio connection, face flat image and face depth are obtained from existing face database
Image.For another example executive agent can pass through image capture device (such as image capture device shown in FIG. 1 1011,1012)
The facial image for acquiring target object, to obtain face flat image and face depth image.
Herein, face flat image typically refers to two-dimension human face image, can be coloured image with (but not limited to), such as
RGB (Red, Green, Blue, RGB) photo.Face depth image typically refers to include the image of face depth information.Figure
The specific format of picture is not intended to limit in this application, such as jpg (Joint Photo graphic Experts Group, Yi Zhongtu
Piece format), the formats such as BMP (Bitmap, image file format) or RAW (RAW Image Format, nondestructive compression type), only
It wants that main body identification can be performed.
In the present embodiment, target object can be arbitrary user, such as using terminal (such as terminal 101 shown in FIG. 1)
User, appear in the user etc. in the acquisition range of image capture device for another example.
As an example, executive agent can be by the face of target object that image capture device is acquired under any attitude
Image is as face flat image.And executive agent can be according to the location information of image capture device and the mesh acquired
Each facial image for marking object, determines the face depth data of target object, to generate the face depth image of target object.
Wherein, the human face posture of face depth image can be identical as the human face posture of face flat image with (but not limited to).Face is deep
The human face posture of degree image can also be the posture in any facial image for the target object that image capture device is acquired.Its
In, posture refers generally to appearance expression of the target object under different visual angles.Here posture is primarily referred to as visual angle, i.e. face
Rotation angle relative to same datum mark.
The present embodiment some optionally in realization method, face depth image can be obtained by following steps：It holds
Row main body can receive the face depth image for the target object that depth image collecting device is acquired；Or receive at least two
The facial image for the target object that image capture device acquires simultaneously；Then according to the facial image of reception, target can be generated
The face depth image of object.Here regular hour deviation is allowed for while.
As an example, when terminal is towards when being provided only with camera (a such as front camera) on the side of user, it should
Camera can be with time-sharing multiplex.I.e. the camera both can be used for acquiring the face flat image of user, at this time commonly to image
Head；The face depth image that can be used for acquisition user, is at this time depth camera.Since face flat image and face are deep
Degree image is usually to be acquired within the similar time, it is possible to think the face posture having the same in the two.
If there are two front cameras for terminal setting, at this point, user can simultaneously be adopted by the two front cameras
Collect two facial images.It is understood that the two front cameras would be normally located at different positions, so respectively
The angle (posture) of the facial image of acquisition is different.Therefore, according to the two of acquisition facial images, it is preposition that the two can be obtained
Face depth image in camera under the acquisition angles of any front camera, i.e., the face depth map of two kinds different postures
Picture.
It should be noted that above-mentioned facial image can be acquired (such as to take pictures, record to the real human face of target object
As etc.) obtain, can also be to be acquired to the carrier of the facial image comprising target object.Wherein, carrier can
To include the screen of photo in kind and/or electronic equipment.That is, by acquire target object human face photo or screen on
Face electronic photo, video or simulation animation of the target object of presentation etc. can also obtain the facial image of target object.It can
With understanding, the facial image and the real human face that are presented on photo or screen are different, the former face is plane, then
The face of person is three-dimensional.
At this point, in order to improve analysis efficiency, executive agent after receiving the facial image that image capture device is acquired,
It can determine whether the face for being currently located at the target object before image capture device is three-dimensional according to these facial images first
Structure；Later if it is determined that the face of target object is stereochemical structure, then the face depth image of target object can be generated.
If that is, the object (real human face of target object or comprising target pair acquired when image capture device
The carrier of the facial image of elephant) it is stereochemical structure, illustrate that the object is likely to be real human face, then can generate corresponding face
Depth image is further analyzed to be done to the face depth image.And when the object that image capture device is acquired is flat
When the structure of face, illustrate the object certainly not real human face.Hence, it can be determined that human face analysis result is failure or does not pass through.
It is not necessarily to human-computer interaction (do not need user's cooperation and do required movement) in this way, so that it may with what is presented to the screen of photo and electronic equipment
Facial image is effectively detected, and so as to avoid some illegal acts, reduces the property loss risk of target object.
Optionally, above-mentioned face flat image may include following at least one：Near-infrared image, gray level image and colour
Image.Correspondingly, above-mentioned image capture device can be near-infrared camera or common camera etc..
Step 202, according to face flat image, simulation face depth data of the generation target object in preset posture.
In the present embodiment, according to the face flat image obtained in step 201, executive agent can be to target object
Human face structure is analyzed, to generate target object preset posture simulation face depth data.Wherein, simulation face is deep
Degrees of data approximate can truly describe face depth data of the real human face of target object under preset posture.Here pre-
If posture is not intended to limit in this application, can be arranged according to actual conditions.Under normal conditions, preset posture is the people obtained
Human face posture in face depth image.
The present embodiment some optionally in realization method, face flat image can be inputted advance structure by executive agent
The three-dimensional deformation model (3D Morphable Model, 3DMM) built, so as to the average people in analyzing three-dimensional deformation model
The deviation between face in face and face flat image.In this way, according to these deviations, the feature ginseng of average face can be adjusted
Number, so as to generate target object preset posture three-dimensional face model.Here the three-dimensional face model generated can be close
It is similar to the real human face structure of target object.Later according to three-dimensional face model, executive agent can extract target object pre-
If the face depth data under posture, and then generate simulation face depth data.
It is understood that in order to further increase analysis efficiency, the face flat image in three-dimensional deformation model is inputted
Can be image of the target object under preset posture, so as to avoid the three-dimensional face model to generation from carrying out human face posture
Adjustment.Face in the face flat image obtained and the face posture having the same in the face depth image of acquisition.
In addition, the face flat image of input can be as more as possible, the higher image of clarity comprising face characteristic, such as front is without screening
Keep off image, to the left 20 ° of (or to the right) to the side image etc. within the scope of 45 °.
Three-dimensional deformation model has typically represented an average face, but also contains simultaneously and common inclined of this average value
Poor pattern information.For example, the face for such as having grown a long-snouted may also be there are one very long chin.In view of this correlation, hold
Row main body can only list hundreds of number word descriptions in the case where all storing your related face's whole characterization information
The deviation of your face and average face, it will be able to which generation is specific to your face-image.Moreover, these deviations further include
The parameters such as rough age, gender and face's length.It should be noted that three-dimensional deformation model is as common three-dimensional face weight
Technology is built, has been widely used for the fields such as medical treatment, education, amusement at present.This model is commercially available, and details are not described herein.
Optionally, executive agent can also input face flat image in convolutional neural networks trained in advance, to
By the convolutional neural networks can reconstruct target object preset posture three-dimensional face model.Basis reconstructs later
Three-dimensional face model, simulation face depth data of the target object under preset posture can be generated.Wherein, convolutional Neural
Network can be used for reconstructing three-dimensional face model according to face flat image.It can be obtained by following training process with (but not limited to)
To convolutional neural networks.
As an example, can three-dimensional deformation model be inputted the different face flat images of same sample object, to deserved
To multiple three-dimensional face models of the sample object under preset posture.It later, can be to the feature of these three-dimensional face models
Parameter is handled and (such as polymerize or weight), to finally obtain three-dimensional face mould of the sample object under preset posture
Type.It then, can be using the different face flat images of different sample objects as input, by different sample objects in preset posture
Under three-dimensional face model as output, training obtain convolutional neural networks.
Step 203, simulation face depth data and face depth image are subjected to the matching analysis.
In the present embodiment, executive agent can will be in the simulation face depth data that generated in step 202 and step 201
The face depth image of acquisition carries out the matching analysis.Such as common recognition of face analysis method may be used, it such as calculates identical
The similarity of feature (mouth, nose, chin etc.).
It is understood that since real human face has stereoscopic face structure, so ought face be acquired from different perspectives
When image, occlusion area, shaded area for being formed in image etc. also can be different, especially nose areas.So simulation face is deep
Degrees of data can embody the stereochemical structure of face.And it simulates face depth data and is in identical preset with face depth image
Under posture, so can also the data such as the contour area of same characteristic features, blocking position, shadow positions and area be carried out with matching point
Analysis, so that it is determined that whether the face depth image obtained meets the stereoscopic face structure under preset posture.
Optionally, executive agent can also will simulate face depth data and face depth image inputs net trained in advance
Network model, to carry out the matching analysis to the two by the network model.Wherein, network model can be used for identical appearance
The simulation face depth data and face depth image of state carry out the matching analysis, and generate the matching analysis result.It herein, can be with
A large amount of positive negative sample is collected, which is obtained with training.Wherein, positive sample refers mainly to real human face or approximate true people
The sample face depth image of face (such as 3D printing model).And negative sample refers mainly to the sample of non-genuine face (such as the photo of bending)
This face depth image.
Step 204, according to the matching analysis as a result, generating the human face analysis result of target object.
In the present embodiment, according to the matching analysis in step 203 as a result, executive agent can generate the people of target object
Face analysis result.Wherein, human face analysis result can serve to indicate that whether the human face analysis of target object passes through.
The present embodiment some optionally in realization method, first, executive agent can determine the matching analysis the result is that
It is no to meet preset condition.Here preset condition is not intended to limit in this application, can be according to the matching analysis side in step 203
Method is configured, as similarity not less than 90% or error be not more than 5% etc..It is preset if it is determined that the matching analysis result meets
Condition, then executive agent can generate point for meeting stereoscopic face structure under preset posture for characterizing the face of target object
Analyse result.If it is determined that the matching analysis result is unsatisfactory for preset condition, then executive agent can be generated for characterizing target object
Face does not meet the analysis result of the stereoscopic face structure under preset posture.
It is understood that the matching analysis is carried out by the face depth image that will simulate face depth data and acquisition,
In this way without human-computer interaction, although being three-dimensional for those, the image of stereoscopic face structure is not met, with
Although and meet stereoscopic face structure, do not meet the image of the stereoscopic face structure under preset posture, as bend after photo
Or the facial image that Curved screen is presented, effectively analysis and identification can be carried out.That is, if human face analysis result is logical
(i.e. the face of target object meets the stereoscopic face structure under preset posture) is crossed, it can be said that the face depth image of bright acquisition
From real human face.Therefore, the method for analyzing facial image in the present embodiment can be applied to the In vivo detection of face.
In application scenes, executive agent can be combined with other detection methods (can set according to actual conditions
Set), to further determine that whether the facial image of target object comes from live body.For example, continuous or compartment of terrain obtains multiple people
The facial image in these face flat images can be compared in face flat image, executive agent.It is tied in this way according to comparison
Fruit (two states are opened and closed to such as eyes, nozzle type changes), without human-computer interaction, it may be determined that
Whether the facial image of target object comes from live body.If it is determined that coming from live body, then the face point of target object can be determined
Analysis passes through.For another example the terminal that executive agent can be used to target object sends (voice and/or text) action command.It is logical
The action indicated by target object execution action command is crossed, executive agent can determine whether the facial image of target object comes from
In live body, so that it is determined that whether the human face analysis of target object passes through.It in this way can be to the facial image comprising target object
3D printing model, bending the non-living bodies such as photo carry out that analysis is effectively detected, so as to further increase being applicable in for detection
Range and accuracy.
It is an application scenarios according to the method for analyzing facial image of the present embodiment with further reference to Fig. 3, Fig. 3
Schematic diagram.In the application scenarios of Fig. 3, when user is paid the bill using the payment application installed in terminal 101, terminal
Image capture device on 101 is activated opening.Image capture device can acquire the front face flat image A of user in this way
With front face depth image B.For the ease of user's operation, institute's the image collected can in gatherer process for image capture device
To be shown on the screen of terminal 101.Meanwhile terminal 101 can be by network, by front face flat image A and positive dough figurine
Face depth image B is sent to service 103.
Server 103 can generate the front simulation face depth data C of user first according to front face flat image A.
Later, can front simulation face depth data C and front face depth image B be subjected to the matching analysis, so as to exclude
It is detected analysis using user's facial image of human face photo and screen presentation after the human face photo of user or bending
Situation.(the front face depth image of the user acquired meets under frontal pose after the Face datection analysis of user passes through
Stereoscopic face structure), server 103 can to terminal 101 send analysis result.Payment application in this way in terminal 101 can
To complete payment function according to analysis result.
Method provided in this embodiment for analyzing facial image, by obtain target object face flat image and
Face depth image, so as to according to the face flat image of acquisition, generate mould of the target object under identical preset posture
Anthropomorphic face depth data；Later, simulation face depth data can will be generated to match with the face depth image of acquisition
Analysis；Finally, according to the matching analysis as a result, the human face analysis result of target object can be generated.Utilize face flat image
Simulation face depth data is generated, so as to carry out the matching analysis with the face depth image of acquisition, to determine that the face is deep
Spend whether image comes from real human face.Therefore, this method can also be applied to face In vivo detection.
With continued reference to Fig. 4, as the realization to method shown in above-mentioned each figure, this application provides one kind for analyzing face
One embodiment of the device of image.The device embodiment is corresponding with embodiment of the method shown in Fig. 2, which specifically can be with
Applied in various electronic equipments.
As shown in figure 4, the device 400 for analyzing facial image of the present embodiment may include：Acquiring unit 401, matches
Set the face flat image and face depth image for obtaining target object；Data generating unit 402, is configured to according to people
Face flat image, simulation face depth data of the generation target object in preset posture；Analytic unit 403 is configured to mould
Anthropomorphic face depth data carries out the matching analysis with face depth image；As a result generation unit 404 are configured to according to the matching analysis
As a result, generating the human face analysis result of target object.
In some optional realization methods of the present embodiment, data generating unit 402 may include：First generates son list
First (not shown) is configured to face flat image inputting the three-dimensional deformation model built in advance, generates target object
In the three-dimensional face model of preset posture；Second generates subelement (not shown), is configured to according to three-dimensional face model,
Face depth data of the target object under preset posture is extracted, simulation face depth data is generated.
Optionally, as a result generation unit 404 may include：Determination subelement (not shown) is configured to determining
With analysis the result is that no meet preset condition；First response subelement (not shown), is configured to match in response to determining
Analysis result meets preset condition, generates and meets stereoscopic face structure under preset posture for characterizing the face of target object
Analysis result；Second response subelement (not shown), is configured in response to determining that the matching analysis result is unsatisfactory for presetting
Condition generates the analysis result that the stereoscopic face structure under preset posture is not met for characterizing the face of target object.
Further, which can also include image generation unit (not shown), be configured to：It receives deep
The face depth image for the target object that degree image capture device is acquired；Or receive at least two image capture devices simultaneously
The facial image of the target object of acquisition；According to the facial image of reception, the face depth image of target object is generated.
In some embodiments, image generation unit can be further configured to：According to the facial image of reception, determine
Whether the face for being currently located at the target object before image capture device is stereochemical structure；In response to determining the face of target object
For stereochemical structure, the face depth image of target object is generated.
Optionally, face flat image may include (but not limited to) or less at least one：Near-infrared image, gray level image
And coloured image.
It is understood that all units described in the device 400 and each step phase in the method described with reference to figure 2
It is corresponding.As a result, above with respect to the operation of method description, the advantageous effect of feature and generation be equally applicable to device 400 and its
In include unit, details are not described herein.
Referring to Fig. 5, it illustrates the computer systems 500 suitable for the electronic equipment for realizing the embodiment of the present application
Structural schematic diagram.Electronic equipment shown in Fig. 5 is only an example, to the function of the embodiment of the present application and should not use model
Shroud carrys out any restrictions.
As shown in figure 5, computer system 500 includes central processing unit (CPU) 501, it can be read-only according to being stored in
Program in memory (ROM) 502 or be loaded into the program in random access storage device (RAM) 503 from storage section 508 and
Execute various actions appropriate and processing.In RAM 503, also it is stored with system 500 and operates required various programs and data.
CPU 501, ROM 502 and RAM 503 are connected with each other by bus 504.Input/output (I/O) interface 505 is also connected to always
Line 504.
It is connected to I/O interfaces 505 with lower component：Importation 506 including touch screen, keyboard, mouse etc.；Including such as
The output par, c 507 of cathode-ray tube (CRT), liquid crystal display (LCD) etc. and loud speaker etc.；Storage part including hard disk etc.
Divide 508；And the communications portion 509 of the network interface card including LAN card, modem etc..Communications portion 509 via
The network of such as internet executes communication process.Driver 510 is also according to needing to be connected to I/O interfaces 505.Detachable media
511, such as disk, CD, magneto-optic disk, semiconductor memory etc., as needed be mounted on driver 510 on, in order to from
The computer program read thereon is mounted into storage section 508 as needed.
Particularly, in accordance with an embodiment of the present disclosure, it may be implemented as computer above with reference to the process of flow chart description
Software program.For example, embodiment of the disclosure includes a kind of computer program product comprising be carried on computer-readable medium
On computer program, which includes the program code for method shown in execution flow chart.In such reality
It applies in example, which can be downloaded and installed by communications portion 509 from network, and/or from detachable media
511 are mounted.When the computer program is executed by central processing unit (CPU) 501, limited in execution the present processes
Above-mentioned function.It should be noted that the computer-readable medium of the application can be computer-readable signal media or calculating
Machine readable storage medium storing program for executing either the two arbitrarily combines.Computer readable storage medium for example can be --- but it is unlimited
In --- electricity, system, device or the device of magnetic, optical, electromagnetic, infrared ray or semiconductor, or the arbitrary above combination.It calculates
The more specific example of machine readable storage medium storing program for executing can include but is not limited to：Being electrically connected, be portable with one or more conducting wires
Formula computer disk, hard disk, random access storage device (RAM), read-only memory (ROM), erasable programmable read only memory
(EPROM or flash memory), optical fiber, portable compact disc read-only memory (CD-ROM), light storage device, magnetic memory device or
The above-mentioned any appropriate combination of person.In this application, can be any include computer-readable medium or storage program has
Shape medium, the program can be commanded the either device use or in connection of execution system, device.And in the application
In, computer-readable signal media may include in a base band or as the data-signal that a carrier wave part is propagated, wherein
Carry computer-readable program code.Diversified forms may be used in the data-signal of this propagation, including but not limited to electric
Magnetic signal, optical signal or above-mentioned any appropriate combination.Computer-readable signal media can also be computer-readable and deposit
Any computer-readable medium other than storage media, the computer-readable medium can send, propagate or transmit for by referring to
Enable execution system, device either device use or program in connection.The program for including on computer-readable medium
Code can transmit with any suitable medium, including but not limited to：Wirelessly, electric wire, optical cable, RF etc. or above-mentioned times
The suitable combination of meaning.
Flow chart in attached drawing and block diagram, it is illustrated that according to the system of the various embodiments of the application, method and computer journey
The architecture, function and operation in the cards of sequence product.In this regard, each box in flowchart or block diagram can generation
A part for a part for one module, program segment, or code of table, the module, program segment, or code includes one or more uses
The executable instruction of the logic function as defined in realization.It should also be noted that in some implementations as replacements, being marked in box
The function of note can also occur in a different order than that indicated in the drawings.For example, two boxes succeedingly indicated are actually
It can be basically executed in parallel, they can also be executed in the opposite order sometimes, this is depended on the functions involved.Also it to note
Meaning, the combination of each box in block diagram and or flow chart and the box in block diagram and or flow chart can be with holding
The dedicated hardware based system of functions or operations as defined in row is realized, or can use specialized hardware and computer instruction
Combination realize.
Being described in unit involved in the embodiment of the present application can be realized by way of software, can also be by hard
The mode of part is realized.Described unit can also be arranged in the processor, for example, can be described as：A kind of processor packet
Include acquiring unit, data generating unit, analytic unit and result generation unit.Wherein, the title of these units is in certain situation
Under do not constitute restriction to the unit itself, for example, acquiring unit is also described as, " face for obtaining target object is flat
The unit of face image and face depth image ".
As on the other hand, present invention also provides a kind of computer-readable medium, which can be
Included in electronic equipment described in above-described embodiment；Can also be individualism, and without be incorporated the electronic equipment in.
Above computer readable medium carries one or more program, when said one or multiple programs are held by the electronic equipment
When row so that the electronic equipment：Obtain the face flat image and face depth image of target object；According to face plan view
Picture, simulation face depth data of the generation target object in preset posture；It will simulation face depth data and face depth image
Carry out the matching analysis；According to the matching analysis as a result, generating the human face analysis result of target object.
Above description is only the preferred embodiment of the application and the explanation to institute's application technology principle.People in the art
Member should be appreciated that invention scope involved in the application, however it is not limited to technology made of the specific combination of above-mentioned technical characteristic
Scheme, while should also cover in the case where not departing from foregoing invention design, it is carried out by above-mentioned technical characteristic or its equivalent feature
Other technical solutions of arbitrary combination and formation.Such as features described above has similar work(with (but not limited to) disclosed herein
Can technical characteristic replaced mutually and the technical solution that is formed.
Claims (14)
1. a kind of method for analyzing facial image, including：
Obtain the face depth image of the face flat image sum of target object；
According to the face flat image, generate the target object the preset posture simulation face depth data；
The simulation face depth data and the face depth image are subjected to the matching analysis；
According to the matching analysis as a result, generating the human face analysis result of the target object.
2. it is described according to the face flat image according to the method described in claim 1, wherein, generate the target object
In the simulation face depth data of the preset posture, including：
The face flat image is inputted to the three-dimensional deformation model built in advance, generates the target object in the default appearance
The three-dimensional face model of state；
According to the three-dimensional face model, face depth data of the target object under the preset posture is extracted, is generated
Simulate face depth data.
3. according to the method described in claim 1, wherein, it is described according to the matching analysis as a result, generating the people of the target object
Face analysis is as a result, include：
Determine whether the matching analysis result meets preset condition；
In response to determining that the matching analysis result meets preset condition, generate for characterize the face of the target object meet it is described
The analysis result of stereoscopic face structure under preset posture；
In response to determining that the matching analysis result is unsatisfactory for preset condition, generation is not met for characterizing the face of the target object
The analysis result of stereoscopic face structure under the preset posture.
4. according to the method described in claim 1, wherein, the face depth image is obtained by following steps：
Receive the face depth image for the target object that depth image collecting device is acquired；Or
Receive the facial image of at least two image capture devices while the target object of acquisition；According to the face figure of reception
Picture generates the face depth image of the target object.
5. according to the method described in claim 4, wherein, the facial image according to reception generates the target object
Face depth image, including：
According to the facial image of reception, determination is currently located at whether the face of the target object before image capture device is vertical
Body structure；
Face in response to the determination target object is stereochemical structure, generates the face depth image of the target object.
6. according to the method described in one of claim 1-5, wherein the face flat image includes following at least one：Closely
Infrared image, gray level image and coloured image.
7. a kind of device for analyzing facial image, including：
Acquiring unit is configured to obtain the face flat image and face depth image of target object；
Data generating unit is configured to, according to the face flat image, generate the target object in the preset posture
Simulation face depth data；
Analytic unit is configured to the simulation face depth data and the face depth image carrying out the matching analysis；
As a result generation unit is configured to according to the matching analysis as a result, generating the human face analysis result of the target object.
8. device according to claim 7, wherein the data generating unit includes：
First generates subelement, is configured to the face flat image inputting the three-dimensional deformation model built in advance, generate
Three-dimensional face model of the target object in the preset posture；
Second generates subelement, is configured to, according to the three-dimensional face model, extract the target object in the default appearance
Face depth data under state generates simulation face depth data.
9. device according to claim 7, wherein the result generation unit includes：
Determination subelement is configured to determine whether the matching analysis result meets preset condition；
First response subelement is configured to, in response to determining that the matching analysis result meets preset condition, generate for characterizing
The face for stating target object meets the analysis result of the stereoscopic face structure under the preset posture；
Second response subelement is configured to, in response to determining that the matching analysis result is unsatisfactory for preset condition, generate for characterizing
The face of the target object does not meet the analysis result of the stereoscopic face structure under the preset posture.
10. device according to claim 7, wherein described device further includes image generation unit, is configured to：
Receive the face depth image for the target object that depth image collecting device is acquired；Or
Receive the facial image of at least two image capture devices while the target object of acquisition；According to the face figure of reception
Picture generates the face depth image of the target object.
11. device according to claim 10, wherein described image generation unit is further configured to：
According to the facial image of reception, determination is currently located at whether the face of the target object before image capture device is vertical
Body structure；
Face in response to the determination target object is stereochemical structure, generates the face depth image of the target object.
12. according to the device described in one of claim 7-11, wherein the face flat image includes following at least one：
Near-infrared image, gray level image and coloured image.
13. a kind of electronic equipment, including：
One or more processors；
Storage device, for storing one or more programs；
When one or more of programs are executed by one or more of processors so that one or more of processors are real
The now method as described in any in claim 1-6.
14. a kind of computer-readable medium, is stored thereon with computer program, wherein the computer program is held by processor
The method as described in any in claim 1-6 is realized when row.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810247286.1A CN108388889B (en) | 2018-03-23 | 2018-03-23 | Method and device for analyzing face image |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810247286.1A CN108388889B (en) | 2018-03-23 | 2018-03-23 | Method and device for analyzing face image |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108388889A true CN108388889A (en) | 2018-08-10 |
| CN108388889B CN108388889B (en) | 2022-02-18 |
Family
ID=63068134
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810247286.1A Active CN108388889B (en) | 2018-03-23 | 2018-03-23 | Method and device for analyzing face image |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108388889B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109101871A (en) * | 2018-08-07 | 2018-12-28 | 北京华捷艾米科技有限公司 | A kind of living body detection device based on depth and Near Infrared Information, detection method and its application |
| CN109542230A (en) * | 2018-11-28 | 2019-03-29 | 北京旷视科技有限公司 | Image processing method, device, electronic equipment and storage medium |
| WO2020037679A1 (en) * | 2018-08-24 | 2020-02-27 | 太平洋未来科技（深圳）有限公司 | Video processing method and apparatus, and electronic device |
| WO2020037678A1 (en) * | 2018-08-24 | 2020-02-27 | 太平洋未来科技（深圳）有限公司 | Method, device, and electronic apparatus for generating three-dimensional human face image from occluded image |
| WO2020088588A1 (en) * | 2018-11-01 | 2020-05-07 | 长沙小钴科技有限公司 | Deep learning-based static three-dimensional method for detecting whether face belongs to living body |
Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101999900A (en) * | 2009-08-28 | 2011-04-06 | 南京壹进制信息技术有限公司 | Living body detecting method and system applied to human face recognition |
| US20150154392A1 (en) * | 2013-11-29 | 2015-06-04 | International Business Machines Corporation | Secure face authentication with liveness detection for mobile |
| CN104834901A (en) * | 2015-04-17 | 2015-08-12 | 北京海鑫科金高科技股份有限公司 | Binocular stereo vision-based human face detection method, device and system |
| CN105022994A (en) * | 2015-06-30 | 2015-11-04 | 国网山东省电力公司日照供电公司 | Identity authentication method of network safety access of power system |
| CN105654048A (en) * | 2015-12-30 | 2016-06-08 | 四川川大智胜软件股份有限公司 | Multi-visual-angle face comparison method |
| EP3061023A1 (en) * | 2013-10-25 | 2016-08-31 | Christophe Remillet | A method and a system for performing 3d-based identity verification of individuals with mobile devices |
| CN106874871A (en) * | 2017-02-15 | 2017-06-20 | 广东光阵光电科技有限公司 | A kind of recognition methods of living body faces dual camera and identifying device |
| CN107609383A (en) * | 2017-10-26 | 2018-01-19 | 深圳奥比中光科技有限公司 | 3D face identity authentications and device |
-
2018
- 2018-03-23 CN CN201810247286.1A patent/CN108388889B/en active Active
Patent Citations (8)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN101999900A (en) * | 2009-08-28 | 2011-04-06 | 南京壹进制信息技术有限公司 | Living body detecting method and system applied to human face recognition |
| EP3061023A1 (en) * | 2013-10-25 | 2016-08-31 | Christophe Remillet | A method and a system for performing 3d-based identity verification of individuals with mobile devices |
| US20150154392A1 (en) * | 2013-11-29 | 2015-06-04 | International Business Machines Corporation | Secure face authentication with liveness detection for mobile |
| CN104834901A (en) * | 2015-04-17 | 2015-08-12 | 北京海鑫科金高科技股份有限公司 | Binocular stereo vision-based human face detection method, device and system |
| CN105022994A (en) * | 2015-06-30 | 2015-11-04 | 国网山东省电力公司日照供电公司 | Identity authentication method of network safety access of power system |
| CN105654048A (en) * | 2015-12-30 | 2016-06-08 | 四川川大智胜软件股份有限公司 | Multi-visual-angle face comparison method |
| CN106874871A (en) * | 2017-02-15 | 2017-06-20 | 广东光阵光电科技有限公司 | A kind of recognition methods of living body faces dual camera and identifying device |
| CN107609383A (en) * | 2017-10-26 | 2018-01-19 | 深圳奥比中光科技有限公司 | 3D face identity authentications and device |
Cited By (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109101871A (en) * | 2018-08-07 | 2018-12-28 | 北京华捷艾米科技有限公司 | A kind of living body detection device based on depth and Near Infrared Information, detection method and its application |
| WO2020037679A1 (en) * | 2018-08-24 | 2020-02-27 | 太平洋未来科技（深圳）有限公司 | Video processing method and apparatus, and electronic device |
| WO2020037678A1 (en) * | 2018-08-24 | 2020-02-27 | 太平洋未来科技（深圳）有限公司 | Method, device, and electronic apparatus for generating three-dimensional human face image from occluded image |
| WO2020088588A1 (en) * | 2018-11-01 | 2020-05-07 | 长沙小钴科技有限公司 | Deep learning-based static three-dimensional method for detecting whether face belongs to living body |
| CN109542230A (en) * | 2018-11-28 | 2019-03-29 | 北京旷视科技有限公司 | Image processing method, device, electronic equipment and storage medium |
| CN109542230B (en) * | 2018-11-28 | 2022-09-27 | 北京旷视科技有限公司 | Image processing method, image processing device, electronic equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108388889B (en) | 2022-02-18 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US11010896B2 (en) | Methods and systems for generating 3D datasets to train deep learning networks for measurements estimation | |
| CN108961369B (en) | Method and device for generating 3D animation | |
| EP3885965B1 (en) | Image recognition method based on micro facial expressions, apparatus and related device | |
| CN109902659B (en) | Method and apparatus for processing human body image | |
| CN108537152A (en) | Method and apparatus for detecting live body | |
| CN108388889A (en) | Method and apparatus for analyzing facial image | |
| CN108416324A (en) | Method and apparatus for detecting live body | |
| CN108509915A (en) | The generation method and device of human face recognition model | |
| CN110210393A (en) | The detection method and device of facial image | |
| CN108898185A (en) | Method and apparatus for generating image recognition model | |
| CN108363995A (en) | Method and apparatus for generating data | |
| CN109255830A (en) | Three-dimensional facial reconstruction method and device | |
| CN107578017A (en) | Method and apparatus for generating image | |
| CN108776786A (en) | Method and apparatus for generating user's truth identification model | |
| CN110298319B (en) | Image synthesis method and device | |
| CN110245612A (en) | The detection method and device of facial image | |
| CN113822977A (en) | Image rendering method, device, equipment and storage medium | |
| CN108197618A (en) | For generating the method and apparatus of Face datection model | |
| CN109086719A (en) | Method and apparatus for output data | |
| CN109887077B (en) | Method and apparatus for generating three-dimensional model | |
| CN108229375B (en) | Method and device for detecting face image | |
| CN108491881A (en) | Method and apparatus for generating detection model | |
| CN108491808B (en) | Method and device for acquiring information | |
| CN108062544A (en) | For the method and apparatus of face In vivo detection | |
| CN108509892A (en) | Method and apparatus for generating near-infrared image |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |