CN101908140A - Biopsy method for use in human face identification - Google Patents

Biopsy method for use in human face identification Download PDF

Info

Publication number
CN101908140A
CN101908140A CN2010102405803A CN201010240580A CN101908140A CN 101908140 A CN101908140 A CN 101908140A CN 2010102405803 A CN2010102405803 A CN 2010102405803A CN 201010240580 A CN201010240580 A CN 201010240580A CN 101908140 A CN101908140 A CN 101908140A
Authority
CN
China
Prior art keywords
user
eye
optical flow
biopsy
live
Prior art date
Application number
CN2010102405803A
Other languages
Chinese (zh)
Inventor
马争鸣
李静
刘金葵
谭恒良
Original Assignee
中山大学
Priority date (The priority date 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 date listed.)
Filing date
Publication date
Application filed by 中山大学 filed Critical 中山大学
Priority to CN2010102405803A priority Critical patent/CN101908140A/en
Publication of CN101908140A publication Critical patent/CN101908140A/en

Links

Abstract

The invention provides a biopsy method for use in human face identification and belongs to the technical field of mode identification. The algorithm provided by the invention comprises: firstly, giving a 'please-face-the-camera-with-face' prompt to a user logging in regardless of the current posture of the user, finding the face and eye socket areas by using an Adaboost human face sorter during the posture correction of the user, determining upper and low eyelids and left and right eye corners by using differential projection and precisely framing the positions of the eye sockets; secondly, computing optical flow field of two adjacent frames in an input video sequence by using an LK algorithm; and thirdly, further processing obtained optical flow data to obtain an optical flow amplitude, acquiring the number of the pixels with high amplitude, computing the specific gravity of the pixels with high amplitude, and determining eye movement if the proportion is big. Experiments show that for a real human face, eye movement can be detected easily due to a heavy optical flow generated on the eyes in a posture correction and eye twinkling process, while for a picture, eyes only perform a micro movement no matter how the picture moves or translations. When the method is used for biopsy, a better effect can be obtained.

Description

A kind of biopsy method of in recognition of face, using
Technical field
The invention belongs to area of pattern recognition, be specifically related to a kind of biopsy method of in recognition of face, using based on misleading information and optical flow field.
Background technology
Development along with biological identification technology, it is ripe that face recognition technology has been tending towards, under good illumination condition and attitude, face identification system can carry out people's face preferably and detect and identification, but, such as for the face recognition technology of systems such as gate inhibition, login, the user can rely on illegal means fraud systems such as photo for some.For this type systematic, discrimination is high more, and potential safety hazard is big more.Therefore, live body detects becomes the bottleneck that this type systematic is used face recognition technology.
Live body in traditional face identification system detects and is broadly divided into user's cooperation and two kinds of methods of blind Detecting.The method that the user cooperates requires the user to do some specific actions, for example variation of mouth type, head rotation, expression shape change etc. according to the prompting of system.Whether systematic comparison user's reaction is consistent with expection, then is judged as live body if meet, otherwise is judged as photo.The advantage of the detection method that the user cooperates is to guarantee live body, and shortcoming is to have exposed the basis for estimation that system's live body detects.In addition, the detection method that the user cooperates requires the user to do a series of action according to prompting, makes the user produce boredom easily, has reduced the friendly of system.The method of blind Detecting does not need the user to cooperate, and is to carry out live body to detect under the unwitting state of user.The advantage of blind checking method is disguised strong, and is safe, and shortcoming is that can to stablize the live body feature of effective utilization less.For example, it is blind checking method live body feature commonly used that eyes blink, but according to statistics, the eyes that just blinked a time in 2 to 4 seconds for each person, this time is oversize, moreover the people who does not bat an eyelid for a long time is quite a few.
The present invention proposes a kind of people's face biopsy method between user's fitting method and blind checking method.At present, many face identification systems tend to point out the user positive in the face of camera in order to improve discrimination.The biopsy method that the present invention proposes is positive in the face of under the situation of camera the user, still constantly points out the user positive in the face of camera, thereby causes that the user is puzzled and subconsciously adjust oneself attitude.With regard to normal user, the action of these adjustment can be expected.The biopsy method that the present invention proposes is judged live body by detecting these actions.
1981, and Horn and Schunck (list of references [1]: B.K.P.Horn, B.G.Schunck, DeterminingOptical Flow, Artificial Intelligence 17,1981 pp.185-203) studies the calculating of optical flow field the earliest.Light stream refers to grayscale mode movement velocity in the image.Object is under light source irradiation, and its surperficial gray scale presents certain space distribution, is referred to as grayscale mode.When people's eyes are observed moving object, the scene of object forms a series of continually varying images on the retina of human eye, these a series of continually varying information are " flowing through " retina (being the plane of delineation) constantly, seems a kind of " stream " of light, so be referred to as light stream.When object of which movement, the luminance patterns of corresponding object is also in motion on image.Light stream is meant apparent (or apparent) motion (apparent motion) of brightness of image pattern.The variation of image is expressed in light stream, comprises the information of target travel, can be used to determine the motion of target.The definition light stream is based on point.Specifically, establish that (u v) is that (x, light stream y) is then (u v) becomes the light stream point for x, y for picture point.The set of all light stream points is called optical flow field.The present invention utilizes the unified interior eyes of eye socket that detect of optical flow field calculating to blink and oculogyral action.
Summary of the invention
The present invention proposes a kind of recognition of face biopsy method based on misleading information and optical flow field, and particular content is as follows:
(1) system provides misleading information
At present, many face identification systems all can point out the user positive in the face of camera in order to improve discrimination.The biopsy method that the present invention proposes is in the positive information in the face of constantly prompting still " positive towards camera " under the situation of camera of user, thereby makes the user produce the puzzled and subconscious attitude of adjusting oneself.System carries out the live body judgement by detecting these attitudes.Done three purposes like this:
Guarantee the disguise that live body detects.Information of the present invention all is an information that many face identification systems often use in " application that the user cooperates ", normal, as " please be positive in the face of camera ", " on-the-spot light is dark excessively ", " please extract glasses " or the like.These information originally are not for live body detects, but in order to improve discrimination.These information are extensive use of in " application that the user cooperates " and widely are familiar with.The user sees that these information can not become suspicious, and can not know that system is carrying out live body and detecting, thereby guarantee the disguise that live body detects.
Lure that the user produces the live body action into.The present invention has satisfied under the prerequisite of information still display reminding information constantly the user.For example, positive the user in the face of under the situation of camera, still point out the user positive in the face of camera.In the present invention, we call misleading information to these information.The purpose of misleading information is to bring out the user to produce puzzlement, thereby subconsciously adjusts the attitude of oneself.Different users may have different attitude adjustment actions.For example, can swing oneself head of the user who has, the user who has can swing up and down the head of oneself, or the like.But under the strong psychological hint of " positive in the face of camera ", the user is the time marquis who adjusts own attitude, and eyes can be kept a close watch on display screen always.This is because user's the video and the information of system all are presented on the display screen in real time, and the user need grasp the effect that own attitude is adjusted in real time by the image and the information of observing on the display screen.The eyeball that this means the user can change when user's attitude is adjusted at the relative position of its eye socket.The present invention judges live body by detecting eyeball with respect to the change in location of eye socket.
Impel the user to use kopiopia, bring out action nictation.Be that live body detects distinguishing rule commonly used in the recognition of face nictation.But according to statistics, human eye just blinked once in average 2 to 4 seconds, and the people who does not bat an eyelid for a long time is quite a few.Therefore, bringing out the user, to produce nictation action under unconscious state be the problem of blind checking method overriding concern.The present invention adopts misleading information to impel the user to produce puzzlement, adopts the mode have good luck or to flash constantly to show these misleading information again, causes user's spirit and with the fatigue of eye, brings out the user and blink.The present invention is by detecting user's judgement live body of blinking.
(2) Adaboost detects people's face, human eye area
At first image is carried out certain light compensation, utilize Adaboost cascade type people face sorter that input picture is carried out people's face then and detect, utilize Adaboost cascade type eyes sorter to carry out human eye detection at the detected human face region first half again.The detected human eye eye socket of Adaboost has comprised some skin areas outside eyebrow zone, the canthus, the eyes left and right sides and a part of skin region of lower part of eye, therefore need eyebrow be excised according to fixed proportion, obtains candidate's eye areas.
(3) the difference Gray Projection is accurately confined the human eye eye socket
For accurately confining eye socket, in candidate's eye areas, carry out the difference Gray Projection.Difference Gray Projection function has reacted the grey scale change size of image on certain direction.Because there is bigger variation in eyes with respect to the surrounding skin gray-scale value, so after passing through horizontal projection, vertical projection respectively, the relative surrounding skin of the projection value of last lower eyelid, position, canthus, left and right sides zone is big a lot, and characteristic is determined the human eye eye socket thus.
(4) calculate adjacent two frame eyes optical flow fields
For adjacent two frame people face figure, can obtain two width of cloth left eye eye patterns and two width of cloth right eye eye patterns respectively by above step, take the LK algorithm that left eye, right eye are done optical flow computation respectively, purpose is to judge whether eye bigger motion has taken place.For people's face, in the process of attitude correction, can follow goggle and eyes blink, usually need accurately to determine the eyeball position at the algorithm that goggles, but can bring bigger calculated amount like this, and present feature point extraction algorithm robustness deficiency changes responsive for light, attitude; The algorithm that blinks at eyes needs usually to determine that eyes are opened, closure state, then judges activity nictation has taken place if status switch meets the human eye process of blinking.The present invention introduces the notion of optical flow field, can estimate this two kinds of movable informations simultaneously, the calculating effect is improved greatly, the optical flow field that calculates when above-mentioned motion takes place is bigger, for photo, although the motion of photo also can bring certain light stream to change, compared to above-mentioned two kinds of motions of real human face, its optical flow field is faint many.
(5) living body determination
Can obtain the pixel motion amplitude by the result of optical flow computation, calculate enough big the counting of amplitude and account for the proportion that entire image is counted, judge that then this moment, bigger motion took place human eye area, be illustrated as live body, otherwise be photo if account for sizable ratio.
The invention characteristics
One of characteristics of the present invention have been to propose a kind of biopsy method based on the information of misleading.The present invention still constantly points out the user front towards camera in camera in the front the user, so this is a kind of misleading information.The user has a series of reaction in the process of carrying out attitude correction, as rotation head but simultaneously eyes are still watched camera attentively, and at this moment the relative position of eyeball in eye socket will certainly change; Simultaneously, the prompting of information can be brought out the fatigue of eyes of user, makes the easier eyes that blink of user.The present invention has utilized the live body feature of this two aspect to carry out living body determination.This method is a kind of method of discrimination between traditional blind Detecting and user's fit system, belong to blind Detecting in essence, because the user does not also know that system is carrying out living body determination this moment, do not know what kind of live body feature is system utilized yet, when the user proofreaied and correct attitude repeatedly and observes information, he provided system available movable information at just unconscious active.For photo, though its how to rotate, translation, the ocular of facial image is transfixion almost all the time, the optical flow field that is produced is also fainter.One of advantage of this mode is that the live body feature that system extracts is not exposed in face of the user, and can utilize multiple live body feature simultaneously, has strengthened the disguise and the security of system.
Two of characteristics of the present invention are to adopt optical flow field to carry out the estimation of live body feature.The live body feature of utilization of the present invention is that the rotation and the eyes of eyeball blink.For these two kinds motions, at present relevant research algorithm is arranged all, but calculated amount is big and poor effect.The optical flow field theory of utilizing among the present invention can capture simultaneously and goggles and the movable information of eyes when blinking, and does not remove to be concerned about the exact position of eyeball and the concrete state of open and close that eyes blink.The present invention removes the eyebrow part by fixed proportion after Adaboost detects the eye socket approximate region, utilize the method for Gray Projection accurately to confine eye socket then, adopts the optical flow field of LK algorithm computation eye socket position consecutive frame.Write down the light stream amplitude of pixel on every width of cloth figure, the point that the calculating amplitude is enough big accounts for the proportion of being had a few, if this number percent judges then that greater than certain value motion has taken place eye, otherwise static.
Description of drawings
Fig. 1, based on the live body testing process figure of optical flow field
Detected people's face of Fig. 2, Adaboost and eye socket position
Fig. 3, according to the eye socket behind the certain proportion excision Adaboost, obtain candidate's eye socket zone
Fig. 4, Gray Projection figure and eye socket are accurately confined
Embodiment
The present invention detects and the example of discerning that is applied as with the people's face based on video, and specific implementation process of the present invention is described.The plain camera of 500 common everythings is adopted in test, and the distance of camera and people's face is about 0.4 meter, is under the general indoor photoenvironment, and people's face detects, live body detects, recognition system operates on the PC.
Step 1: the misleading information of system prompt
When the user imports username and password expectation and enters system, send an information " please be positive " to the user towards camera, require the user to carry out certain attitude correction, no matter whether the facial image that this moment, whether camera was captured is front view (FV).
Step 2:Adaboost people face, human eye detection
Be introduced into as people's face after the Adaboost algorithm proposes and detect, people's face detection speed is greatly improved.Along with continuous research and the improvement of Chinese scholars, make present Adaboost algorithm become one of main flow algorithm that detects for people's face to its algorithm.
The present invention adopts this algorithm, at first utilizes stackability people face sorter to detect people's face, and to forehead, down to chin, the left and right sides comprises more background image, sees Fig. 2 on detected people's face head portrait; Utilize stackability eyes sorter to carry out searching of eyes in the first half of facial image then, the ocular that finally finds comprises some skin areas of complete eyebrow, canthus, the left and right sides and lower part of eye.Can remove the eyebrow part in proportion, suppose the wide w of the detected human eye eye socket of Adaboost, high h then gives up to the right Give up downwards Remaining As candidate's eye socket, see Fig. 3.This part has tentatively got rid of the eyebrow part, can make next step processing of Gray Projection.
Step 3: the difference Gray Projection is determined the human eye eye socket
Compare with surrounding skin because of human eye, stronger marginal information is arranged, therefore can utilize grey scale change to extract the position at lower eyelid, canthus, the left and right sides.Difference projection has been considered the situation that the rate of gray level of image on certain direction changes, because human eye goes out comparatively remarkable than the grey scale change of surrounding skin, therefore the human eye area of getting rid of eyebrow is carried out difference projection, have a quite tangible peak corresponding to position of human eye.The difference projection function definition is as follows: if picture size is imW * imH, and vertical difference projection DPF v(x) with horizontal difference projection DPF h(y) can be expressed as respectively:
DPF v ( x ) = 1 imW Σ i = 1 imW - 1 | I ( x , y i + 1 ) - I ( x , y i ) |
DPF h ( y ) = 1 imH Σ i = 1 imH - 1 | I ( x i + 1 , y ) - I ( x i , y ) |
The present invention is directed to candidate's eye socket part right and left eyes and carry out the difference projection of level and vertical direction respectively, according to the certain scope of projection function intercepting as accurate eye socket position, specifically as shown in Figure 4.
Step 4: calculate consecutive frame eye socket optical flow field
Each two field picture for the video input extracts its left and right sides eye socket successively, calculates the optical flow field of adjacent two frame correspondences then.Utilize the LK algorithm can obtain the horizontal component image velx and the vertical component image vely of light stream.In order to reduce the system-computed amount, operating rate is improved, the present invention only utilizes central point (x, light stream value y), and these points are become available point of each window window * window.(x y), is averaged the pixel light flow valuve in its velx figure and the vely figure window * window neighborhood respectively, as the horizontal motion components xx and the vertical motion components yy of this point for point.For estimating its movable information v, calculate
The LK algorithm is based on following 3 basic assumptions:
1, the gray consistency of image.Suppose pixel in the image in motion process, its pixel value is changeless, also is that the same point gray-scale value of same object between frame and the frame can not change along with the variation of time;
2, motion amplitude is less.The movement velocity of object is smaller, otherwise the motion conditions less than object is caught in the variation of frame too slowly;
3, consecutive point project behind the 2D plane still adjacent one another arely in the space, have similar kinetic characteristic.
Have by assumed condition, when establishing constantly t, on the image a bit (x, the gray-scale value of y) locating be I (x, y, t).When moment t+ Δ t, this point moves to reposition, and its position on image becomes (x+ Δ x, y+ Δ y), and the gray-scale value of reposition is designated as I (x+ Δ x, y+ Δ y, t+ Δ t), according to gradation of image consistance hypothesis, have dI (x, y, t)/dt=0, then
I(x,y,t)=I(x+Δx,y+Δy,t+Δt)
If u and v are respectively this light stream component in the x and y direction, i.e. u=dx/dt, v=dy/dt launches following formula the right with the Taylor formula, omit the item more than 2 times, then can get the gray level image optical flow field and calculate basic equation
∂ I ∂ x u + ∂ I ∂ y v + ∂ I ∂ t = 0
Promptly
I xu+I yv+I t=0
The Lucas-Kanade method is a kind of optical flow algorithm based on pixel-recursive, predicts that exactly the offset estimation device predicted value of rectification type can be used as the motion estimation value of previous location of pixels, or as the motion estimation linear combination in the current pixel neighborhood.According to the gradient minimum value of the displaced frame difference on this pixel, prediction is done further to revise.
Lucas and Kanade hypothesis Ω motion vector on a little spatial neighborhood keep constant, use weighted least-squares side to estimate light stream then.On a little spatial neighborhood Ω, the light stream evaluated error is defined as
Σ ( x , y ) ∈ Ω W 2 ( x ) ( I x u + I y v + I t ) 2
Wherein W (x) represents the window weighting function, and it makes centre of neighbourhood part bigger than periphery to the influence that constraint produces.
If V=(u, v) T, Separating by following formula of following formula provides:
A TW 2AV=A TW 2b
Wherein, n at moment t puts x i∈ Ω,
A = [ ▿ I ( x 1 ) , . . . ▿ I ( x n ) ] T ,
W=diag[W(x 1),...W(x n)],
b=-(I t(x 1),...I t(x n)) T
A TW 2AV=A TW 2B separates the (A for V= TW 2A) TA TW 2B wherein works as A TW 2A can obtain analytic solution when being nonsingular, because it is the matrix of a 2*2:
A T W 2 A = Σ W 2 ( x ) I x 2 ( x ) Σ W 2 ( x ) I x ( x ) I y ( x ) Σ W 2 ( x ) I y ( x ) I x ( x ) Σ W 2 ( x ) I y 2 ( x ) ,
Wherein all and all be that point on neighborhood Ω obtains.
For the fundamental equation of light stream, the reliability of light stream is by A TW 2The eigenwert of A matrix decides, and supposes that its eigenwert is λ 1And λ 2, if they are all greater than given threshold value λ, the light stream that then calculates is reliably, otherwise, be insecure.
Step 5: judge whether to be live body
For the light stream figure of the left and right sides eye socket that calculates, set a threshold value λ, just it is defined as the point that bigger motion has taken place when having only the motion assignment v of this point 〉=λ.Calculate the enough big point of these motions and occupy the proportion p that effect is counted, if think that eye has produced certain motion during p 〉=15% (number percent can determine that value of the present invention is 15% according to the reality debugging), simultaneously, when having only eyes all to detect movable information, just it is judged to be live body.Experiment shows that for real human face, along with the correction of attitude and blinking of eyes, the light stream value that calculates is bigger, and the some proportion that the light stream value is bigger is also higher, and system can be judged to it figure that motion has taken place; For photo, no matter how translation of photo, rotation, the motion of the eye that is brought is all fainter, exists than big difference with real human face.

Claims (3)

1. biopsy method of using in recognition of face is characterized in that:
A, utilize regular prompt information to carry out disguised live body to detect;
B, utilize the optical flow field unification to blink and oculogyral detection.
2. according to the right 1 described biopsy method of in recognition of face, using, it is characterized in that described steps A specifically comprises: positive in the face of under the situation of camera the user, still constantly the flash information of " please positive in the face of camera ", thereby cause that the user is puzzled, produce the action that can expect that some have the live body feature: understand subconscious rectification attitude and eyes are kept a close watch on screen as the user, this is equivalent to eyeball and has taken place to move in eye socket; Owing to excessively watch information attentively, make eye fatigue and blink.
3. according to the right 1 described biopsy method of in recognition of face, using, it is characterized in that described step B specifically comprises: utilize and calculate the live body action that the unified detection of optical flow field eyeball moves and blinks, because moving or blink, eyeball can unify to be summarised as that object moves in the eye socket, optical flow field is relatively responsive to object of which movement, utilizes optical flow field can unify to detect eyeball and moves and blink.
CN2010102405803A 2010-07-29 2010-07-29 Biopsy method for use in human face identification CN101908140A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
CN2010102405803A CN101908140A (en) 2010-07-29 2010-07-29 Biopsy method for use in human face identification

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
CN2010102405803A CN101908140A (en) 2010-07-29 2010-07-29 Biopsy method for use in human face identification

Publications (1)

Publication Number Publication Date
CN101908140A true CN101908140A (en) 2010-12-08

Family

ID=43263595

Family Applications (1)

Application Number Title Priority Date Filing Date
CN2010102405803A CN101908140A (en) 2010-07-29 2010-07-29 Biopsy method for use in human face identification

Country Status (1)

Country Link
CN (1) CN101908140A (en)

Cited By (31)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103279260A (en) * 2013-04-10 2013-09-04 苏州三星电子电脑有限公司 Direction self-adaptation display system and adjusting method thereof
CN103324879A (en) * 2013-07-05 2013-09-25 公安部第三研究所 System and method for identification verification on mobile terminal and based on face recognition and intelligent card
CN103379282A (en) * 2012-04-26 2013-10-30 三星电子株式会社 Apparatus and method for recognizing image
CN103440479A (en) * 2013-08-29 2013-12-11 湖北微模式科技发展有限公司 Method and system for detecting living body human face
CN103514439A (en) * 2012-06-26 2014-01-15 谷歌公司 Facial recognition
CN103797752A (en) * 2011-06-03 2014-05-14 阿维米尔Ip有限责任公司 Method and computer program for providing authentication to control access to a computer system
CN103927012A (en) * 2014-04-16 2014-07-16 北京尚德智产投资管理有限公司 Mobile terminal achieving user posture detection through operating system and method
CN103955273A (en) * 2014-04-16 2014-07-30 北京尚德智产投资管理有限公司 Mobile terminal and method for realizing user posture detection through operation system
CN104348778A (en) * 2013-07-25 2015-02-11 信帧电子技术(北京)有限公司 Remote identity authentication system, terminal and method carrying out initial face identification at handset terminal
CN104361326A (en) * 2014-11-18 2015-02-18 新开普电子股份有限公司 Method for distinguishing living human face
CN104504378A (en) * 2014-12-29 2015-04-08 北京奇艺世纪科技有限公司 Method and device for detecting image information
CN104751110A (en) * 2013-12-31 2015-07-01 汉王科技股份有限公司 Bio-assay detection method and device
CN105279764A (en) * 2014-05-27 2016-01-27 北京三星通信技术研究有限公司 Eye image processing device and eye image processing method
CN105868677A (en) * 2015-01-19 2016-08-17 阿里巴巴集团控股有限公司 Live human face detection method and device
CN106203372A (en) * 2016-07-19 2016-12-07 奇酷互联网络科技(深圳)有限公司 Biopsy methods based on eyes, device and terminal unit
WO2017070971A1 (en) * 2015-10-31 2017-05-04 华为技术有限公司 Facial authentication method and electronic device
CN103955271B (en) * 2014-04-16 2017-06-09 北京智产科技咨询有限公司 A kind of mobile terminal and method that user's attitude detection is realized by operating system
CN107077589A (en) * 2014-11-13 2017-08-18 英特尔公司 Face personation detection in biometric based on image
CN107111743A (en) * 2014-11-13 2017-08-29 英特尔公司 The vital activity tracked using gradual eyelid is detected
CN107273794A (en) * 2017-04-28 2017-10-20 北京建筑大学 Live body discrimination method and device in a kind of face recognition process
CN107330370A (en) * 2017-06-02 2017-11-07 广州视源电子科技股份有限公司 A kind of brow furrows motion detection method and device and vivo identification method and system
CN107563289A (en) * 2017-07-31 2018-01-09 百度在线网络技术(北京)有限公司 A kind of method, apparatus of In vivo detection, equipment and computer-readable storage medium
CN107643826A (en) * 2017-08-28 2018-01-30 天津大学 A kind of unmanned plane man-machine interaction method based on computer vision and deep learning
CN107688781A (en) * 2017-08-22 2018-02-13 北京小米移动软件有限公司 Face identification method and device
CN107886032A (en) * 2016-09-30 2018-04-06 阿里巴巴集团控股有限公司 Terminal device, smart mobile phone, authentication method and system based on face recognition
CN108664956A (en) * 2018-06-25 2018-10-16 北京航空航天大学 Vivo identification method during human vein collecting biological feature information and device
CN108875497A (en) * 2017-10-27 2018-11-23 北京旷视科技有限公司 The method, apparatus and computer storage medium of In vivo detection
CN109508702A (en) * 2018-12-29 2019-03-22 安徽云森物联网科技有限公司 A kind of three-dimensional face biopsy method based on single image acquisition equipment
CN110147458A (en) * 2019-05-24 2019-08-20 涂哲 A kind of photo screening technique, system and electric terminal
CN110223322A (en) * 2019-05-31 2019-09-10 腾讯科技(深圳)有限公司 Image-recognizing method, device, computer equipment and storage medium
US10666859B2 (en) 2018-04-13 2020-05-26 Boe Technology Group Co., Ltd. Self-photographing control method and device, and electronic device

Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN101159016A (en) * 2007-11-26 2008-04-09 清华大学 Living body detecting method and system based on human face physiologic moving
US20090034801A1 (en) * 2007-08-03 2009-02-05 Hammoud Riad I System and method of awareness detection

Patent Citations (2)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US20090034801A1 (en) * 2007-08-03 2009-02-05 Hammoud Riad I System and method of awareness detection
CN101159016A (en) * 2007-11-26 2008-04-09 清华大学 Living body detecting method and system based on human face physiologic moving

Non-Patent Citations (4)

* Cited by examiner, † Cited by third party
Title
《Fourth IEEE Workshop on Automatic Identification Advanced Technologies, 2005》 20051018 K. Kollreider, et al. Evaluating Liveness by Face Images and the Structure Tensor 第75-80页 1-3 , 2 *
《中国优秀硕士学位论文全文数据库》 20100515 刘江伟 基于人眼检测的驾驶员疲劳状态识别技术 期望 1-3 , 第05期 2 *
《计算机应用研究》 20100430 杨晶东,等 基于Kanade-Lucas的人眼跟踪算法研究 第1575-1577页 1-3 第27卷, 第4期 2 *
《计算机测量与控制》 20071231 曹倩霞,等 基于眼睛特征跟踪的眼睛状态跟踪 第1794页第2栏第2段 1-3 第15卷, 第12期 2 *

Cited By (44)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103797752A (en) * 2011-06-03 2014-05-14 阿维米尔Ip有限责任公司 Method and computer program for providing authentication to control access to a computer system
US9740838B2 (en) 2011-06-03 2017-08-22 Sensipass Ltd. Method and computer program for providing authentication to control access to a computer system
CN103797752B (en) * 2011-06-03 2018-07-17 森西帕斯有限公司 For providing certification to control the access of method and computer to(for) computer system
CN103379282A (en) * 2012-04-26 2013-10-30 三星电子株式会社 Apparatus and method for recognizing image
US9684819B2 (en) 2012-04-26 2017-06-20 Samsung Electronics Co., Ltd. Apparatus and method for distinguishing whether an image is of a live object or a copy of a photo or moving picture
CN103379282B (en) * 2012-04-26 2018-09-21 三星电子株式会社 The device and method of image for identification
US9117109B2 (en) 2012-06-26 2015-08-25 Google Inc. Facial recognition
CN103514439A (en) * 2012-06-26 2014-01-15 谷歌公司 Facial recognition
CN103279260A (en) * 2013-04-10 2013-09-04 苏州三星电子电脑有限公司 Direction self-adaptation display system and adjusting method thereof
CN103324879A (en) * 2013-07-05 2013-09-25 公安部第三研究所 System and method for identification verification on mobile terminal and based on face recognition and intelligent card
CN103324879B (en) * 2013-07-05 2016-08-10 公安部第三研究所 Mobile device is based on recognition of face and the authentication system of smart card and method
CN104348778A (en) * 2013-07-25 2015-02-11 信帧电子技术(北京)有限公司 Remote identity authentication system, terminal and method carrying out initial face identification at handset terminal
CN103440479A (en) * 2013-08-29 2013-12-11 湖北微模式科技发展有限公司 Method and system for detecting living body human face
CN104751110B (en) * 2013-12-31 2018-12-04 汉王科技股份有限公司 A kind of biopsy method and device
CN104751110A (en) * 2013-12-31 2015-07-01 汉王科技股份有限公司 Bio-assay detection method and device
CN103927012B (en) * 2014-04-16 2017-06-13 北京智产科技咨询有限公司 A kind of mobile terminal and method that user's attitude detection is realized by operating system
CN103955273B (en) * 2014-04-16 2017-06-13 北京智产科技咨询有限公司 A kind of mobile terminal and method that user's attitude detection is realized by operating system
CN103955273A (en) * 2014-04-16 2014-07-30 北京尚德智产投资管理有限公司 Mobile terminal and method for realizing user posture detection through operation system
CN103955271B (en) * 2014-04-16 2017-06-09 北京智产科技咨询有限公司 A kind of mobile terminal and method that user's attitude detection is realized by operating system
CN103927012A (en) * 2014-04-16 2014-07-16 北京尚德智产投资管理有限公司 Mobile terminal achieving user posture detection through operating system and method
CN105279764B (en) * 2014-05-27 2020-09-11 北京三星通信技术研究有限公司 Eye image processing apparatus and method
CN105279764A (en) * 2014-05-27 2016-01-27 北京三星通信技术研究有限公司 Eye image processing device and eye image processing method
CN107111743A (en) * 2014-11-13 2017-08-29 英特尔公司 The vital activity tracked using gradual eyelid is detected
CN107077589B (en) * 2014-11-13 2021-02-09 英特尔公司 Facial spoofing detection in image-based biometrics
CN107077589A (en) * 2014-11-13 2017-08-18 英特尔公司 Face personation detection in biometric based on image
CN104361326A (en) * 2014-11-18 2015-02-18 新开普电子股份有限公司 Method for distinguishing living human face
CN104504378A (en) * 2014-12-29 2015-04-08 北京奇艺世纪科技有限公司 Method and device for detecting image information
CN105868677A (en) * 2015-01-19 2016-08-17 阿里巴巴集团控股有限公司 Live human face detection method and device
US10552593B2 (en) 2015-10-31 2020-02-04 Huawei Technologies Co., Ltd. Face verification method and electronic device
WO2017070971A1 (en) * 2015-10-31 2017-05-04 华为技术有限公司 Facial authentication method and electronic device
CN106203372A (en) * 2016-07-19 2016-12-07 奇酷互联网络科技(深圳)有限公司 Biopsy methods based on eyes, device and terminal unit
CN107886032A (en) * 2016-09-30 2018-04-06 阿里巴巴集团控股有限公司 Terminal device, smart mobile phone, authentication method and system based on face recognition
CN107273794A (en) * 2017-04-28 2017-10-20 北京建筑大学 Live body discrimination method and device in a kind of face recognition process
CN107330370B (en) * 2017-06-02 2020-06-19 广州视源电子科技股份有限公司 Forehead wrinkle action detection method and device and living body identification method and system
CN107330370A (en) * 2017-06-02 2017-11-07 广州视源电子科技股份有限公司 A kind of brow furrows motion detection method and device and vivo identification method and system
CN107563289A (en) * 2017-07-31 2018-01-09 百度在线网络技术(北京)有限公司 A kind of method, apparatus of In vivo detection, equipment and computer-readable storage medium
CN107688781A (en) * 2017-08-22 2018-02-13 北京小米移动软件有限公司 Face identification method and device
CN107643826A (en) * 2017-08-28 2018-01-30 天津大学 A kind of unmanned plane man-machine interaction method based on computer vision and deep learning
CN108875497A (en) * 2017-10-27 2018-11-23 北京旷视科技有限公司 The method, apparatus and computer storage medium of In vivo detection
US10666859B2 (en) 2018-04-13 2020-05-26 Boe Technology Group Co., Ltd. Self-photographing control method and device, and electronic device
CN108664956A (en) * 2018-06-25 2018-10-16 北京航空航天大学 Vivo identification method during human vein collecting biological feature information and device
CN109508702A (en) * 2018-12-29 2019-03-22 安徽云森物联网科技有限公司 A kind of three-dimensional face biopsy method based on single image acquisition equipment
CN110147458A (en) * 2019-05-24 2019-08-20 涂哲 A kind of photo screening technique, system and electric terminal
CN110223322A (en) * 2019-05-31 2019-09-10 腾讯科技(深圳)有限公司 Image-recognizing method, device, computer equipment and storage medium

Similar Documents

Publication Publication Date Title
CN106796449B (en) Sight tracking method and device
US9247211B2 (en) System and method for video content analysis using depth sensing
Barra et al. Ubiquitous iris recognition by means of mobile devices
CN104200192B (en) Driver's gaze detection system
CN104013414B (en) A kind of Study in Driver Fatigue State Surveillance System based on intelligent movable mobile phone
Bălan et al. The naked truth: Estimating body shape under clothing
Yu et al. A study on gait-based gender classification
Xiao et al. Robust full‐motion recovery of head by dynamic templates and re‐registration techniques
Grauman et al. Communication via eye blinks and eyebrow raises: Video-based human-computer interfaces
Eriksson et al. Eye-tracking for detection of driver fatigue
Berclaz et al. Robust people tracking with global trajectory optimization
Azim et al. Fully automated real time fatigue detection of drivers through fuzzy expert systems
Tian et al. Dual-state parametric eye tracking
Feng et al. Multi-cues eye detection on gray intensity image
EP1426898B1 (en) Human detection through face detection and motion detection
CN105184246B (en) Living body detection method and living body detection system
US9892316B2 (en) Method and apparatus for pattern tracking
EP2441383B1 (en) Gaze target determination device and gaze target determination method
US6633655B1 (en) Method of and apparatus for detecting a human face and observer tracking display
EP3029604B1 (en) Area information estimating device, area information estimating method, and air conditioning apparatus
Li et al. Learning to predict gaze in egocentric video
CN101593425B (en) Machine vision based fatigue driving monitoring method and system
CN104616438B (en) A kind of motion detection method of yawning for fatigue driving detection
Foroughi et al. Robust fall detection using human shape and multi-class support vector machine
Alioua et al. Driver’s fatigue detection based on yawning extraction

Legal Events

Date Code Title Description
C06 Publication
PB01 Publication
C10 Entry into substantive examination
SE01 Entry into force of request for substantive examination
C02 Deemed withdrawal of patent application after publication (patent law 2001)
WD01 Invention patent application deemed withdrawn after publication

Application publication date: 20101208