CN108363968A - A kind of tired driver driving monitoring system and method based on key point extraction - Google Patents
A kind of tired driver driving monitoring system and method based on key point extraction Download PDFInfo
- Publication number
- CN108363968A CN108363968A CN201810095177.2A CN201810095177A CN108363968A CN 108363968 A CN108363968 A CN 108363968A CN 201810095177 A CN201810095177 A CN 201810095177A CN 108363968 A CN108363968 A CN 108363968A
- Authority
- CN
- China
- Prior art keywords
- face
- driver
- image
- frame
- tired
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V20/00—Scenes; Scene-specific elements
- G06V20/50—Context or environment of the image
- G06V20/59—Context or environment of the image inside of a vehicle, e.g. relating to seat occupancy, driver state or inner lighting conditions
- G06V20/597—Recognising the driver's state or behaviour, e.g. attention or drowsiness
-
- 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
- G06V40/171—Local features and components; Facial parts ; Occluding parts, e.g. glasses; Geometrical relationships
-
- G—PHYSICS
- G08—SIGNALLING
- G08B—SIGNALLING OR CALLING SYSTEMS; ORDER TELEGRAPHS; ALARM SYSTEMS
- G08B21/00—Alarms responsive to a single specified undesired or abnormal condition and not otherwise provided for
- G08B21/02—Alarms for ensuring the safety of persons
- G08B21/06—Alarms for ensuring the safety of persons indicating a condition of sleep, e.g. anti-dozing alarms
Landscapes
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Oral & Maxillofacial Surgery (AREA)
- Physics & Mathematics (AREA)
- General Health & Medical Sciences (AREA)
- Theoretical Computer Science (AREA)
- Multimedia (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Human Computer Interaction (AREA)
- Business, Economics & Management (AREA)
- Emergency Management (AREA)
- Image Processing (AREA)
Abstract
The present invention provides a kind of tired driver driving monitoring system and method extracted based on key point, and the wherein system includes:Image capture module is set as acquiring the face-image of driver in real time;Positioning feature point module is set as positioning the key feature points in every frame face-image;And state determination module, it is set as judging current driver's eyes and the corresponding opening amplitude of face according to the key feature points positioned in each frame face-image, if there is the opening amplitude of the corresponding driver's eyes of Q frame images and face to be less than corresponding predetermined threshold in continuous P frame face-image, then judge that tired driver drives, wherein, P and Q is preset value.The present invention proposes the facial status monitoring means extracted based on key feature points, local feature information when fatigue occurs in driver can be got, it can be with the state of accurate judgement eyes, face, help to improve the accuracy of monitoring, driver is reminded to take a good rest in time, to reduce traffic accident incidence.
Description
Technical field
Monitoring field is driven the present invention relates to tired driver more particularly to a kind of tired driver based on key point extraction is driven
Sail monitoring system and method.
Background technology
Fatigue driving is that cause a big main cause of traffic accident, existing fatigue driving monitoring technology be largely logical
The device of contact is crossed to monitor the physiological characteristic of driver, to judge whether the hazardous act of fatigue driving.However,
The monitoring device of contact can influence the driving behavior of driver to a certain extent, and there are security risks.And it is regarded based on computer
The monitoring method of feel can carry out sentencing for fatigue state in the case where not influencing normal driving using the facial characteristics of driver
It is disconnected.The existing fatigue monitoring method based on image procossing, only gets position of human eye mostly, and there is no for the part such as human eye
Characteristic information further extracts characteristic point.Therefore, facial characteristics how is efficiently extracted out, maximized characteristic information is obtained,
To establish fatigue detecting model, the driving condition of driver is accurately analyzed, is to solve tired driver to drive, reduces traffic accident
One main direction of studying of incidence.
Invention content
The purpose of the present invention is to provide a kind of tired drivers based on key point extraction to drive monitoring system and method, uses
Whether there is fatigue driving during driver driving accurately to monitor, helps to remind driver to take a good rest in time, for ensureing
Traffic safety has important Practical significance.
To achieve the goals above, one aspect of the present invention provides a kind of tired driver driving monitoring extracted based on key point
System, including:
Image capture module is set as acquiring the face-image of driver in real time;
Positioning feature point module is set as positioning the key feature points in every frame face-image;
State determination module is set as judging current driver eye according to the key feature points positioned in each frame face-image
Eyeball and the corresponding opening amplitude of face, if having the corresponding driver's eyes of Q frame images and face in continuous P frame face-image
Opening amplitude is less than corresponding predetermined threshold, then judges that tired driver drives, wherein P and Q is preset value.
Further, described image acquisition module includes camera, infrared light compensating lamp and Video Decoder.
Further, the positioning feature point module is using the key feature points in the every frame face-image of ASM algorithms positioning.
Further, which further includes an image transmission module, and when the state determination module, judgement driver is tired
When please sailing, the face-image that the opening amplitude of a frame driver eyes and face is less than corresponding predetermined threshold is selected to pass through the figure
As transmission module is sent to a host computer.
Further, which further includes an alarm module, when state determination module judgement tired driver is driven
When sailing, the alarm module sends out a fatigue warning signal.
Another aspect of the present invention provides a kind of tired driver driving monitoring method extracted based on key point, and feature exists
In including the following steps:
S1 acquires the face-image of driver;
S2 is positioned per the key feature points in frame face-image;
S3 judges current driver's eyes and the corresponding opening of face according to the key feature points positioned in each frame face-image
Frame degree, if there have the opening amplitude of the corresponding driver's eyes of Q frame images and face to be less than in continuous P frame face-image to be corresponding
Predetermined threshold then judges that tired driver drives, wherein P and Q is preset value.
Further, the step S2 adopts the positioning of ASM algorithms per the key feature points in frame face-image.
Further, the step S3 further includes:When judging that tired driver drives, a frame driver eyes and face are selected
Opening amplitude be less than corresponding predetermined threshold face-image be sent to a host computer.
Further, the step S3 further includes:When state determination module judgement tired driver drives, one is sent out
Fatigue warning signal.
Compared with prior art, the present invention has the advantages that:
The present invention proposes the facial status monitoring means extracted based on key feature points, and it is tired can to get driver's appearance
Local feature information when labor can be helped to improve the accuracy of monitoring, carried in time with the state of accurate judgement eyes, face
Awake driver takes a good rest, to reduce traffic accident incidence.
Description of the drawings
Fig. 1 is the structure diagram that the tired driver extracted the present invention is based on key point drives monitoring system one embodiment.
Specific implementation mode
Above-mentioned and/or additional aspect and advantage of the invention will become from the following description of the accompanying drawings of embodiments
Obviously and it is readily appreciated that.
Fig. 1 shows that the tired driver extracted the present invention is based on key point drives one embodiment of monitoring system, including
Sequentially connected image capture module 1, positioning feature point module 2 and state determination module 3 further include and frame state determination module
The image transmission module 4 and alarm module 5 of 3 connections.Wherein, image capture module 1 includes camera, infrared light compensating lamp and video
Decoder, the face-image for acquiring driver in real time.Positioning feature point module 2 is using existing ASM (active shape model)
Algorithm is used to position the key feature points in every frame face-image, including human eye, face, shape of face, the wheel of the key positions such as nose
Exterior feature orients the characteristic point coordinate come and forms one group of feature vector, to describe the position of key point in a currently processed frame image
Confidence ceases.State determination module 3 be used for according to the key feature points positioned in each frame face-image judge current driver's eyes with
The corresponding opening amplitude of face, if there is the opening of Q frame images corresponding driver's eyes and face in continuous P frame face-image
Amplitude is less than corresponding predetermined threshold, then judges that tired driver drives, wherein P and Q is preset value.When frame state determination module 3
When judging that tired driver drives, the selection wherein opening amplitude of a frame driver eyes and face is less than the face of corresponding predetermined threshold
Image is sent to a host computer by frame image transmission module 4 (such as 4G modules), while command frame alarm module 5 sends out a fatigue
Alarm signal is to prompt driver.
Another aspect of the present invention provides a kind of tired driver driving monitoring method extracted based on key point, including following step
Suddenly:S1 acquires the face-image of driver;S2 is positioned per the key feature points in frame face-image;S3, according to each pattern portion figure
The key feature points positioned as in judge current driver's eyes and the corresponding opening frame degree of face, if in continuous P pattern portion figure
There is the opening amplitude of the corresponding driver's eyes of Q frame images and face to be less than corresponding predetermined threshold as in, then judges tired driver
It drives, when judging that tired driver drives, the opening amplitude of a frame driver eyes and face is selected to be less than corresponding predetermined threshold
Face-image is sent to a host computer and is simultaneously emitted by a fatigue warning signal.
In the present invention, the key feature points in face-image are positioned using ASM algorithms.ASM algorithms include mainly two
A part:First by carrying out characteristic point label to a large amount of facial image sample, n (such as 68) features of face are extracted
Point seeks the points distribution models of face;Then using obtained model, human face region is carried out to the pending image got
Search and matching, gradually approach optimal match point, then it is assumed that the coordinate points finally navigated to are the key of current face's feature
Point, specifically includes following steps:
S21 chooses the N pictures in human face photo library as sample set and carries out n characteristic point to every face-image
Manual calibration, then i-th face-image by n characteristic point coordinate xi=[xi0,yi0,xi1,yi1····xi(n-1),
yi(n-1)]TThe vector of composition indicates that the sample set of N images is expressed as X=[x1,x2,····xN)]T, wherein (xij,
yij) be j-th of characteristic point coordinate, 0≤j < n.
S22 is normalized vectorial X using Procrusts normalization algorithms, to obtain average face model
Then, the covariance matrix S of data is calculated:
Then, the characteristic value and feature vector of covariance matrix are calculated, and characteristic value is arranged into Sp in descending orderi=λipi(λi
≥λi+1;pi Tpi=1;I=1,2 ... ..., N), wherein piIt is the ith feature value λ of SiCorresponding feature vector, extracts S's
Preceding m characteristic value and corresponding feature vector.
S23 carries out dimension-reduction treatment using PCA (principal component analysis) algorithms to X, this step advantageously reduces data volume, carries
High effect.Then construct the statistical shape model of sample:Wherein:P=[P1,P2,P3,P4…Pm] by S before
M feature vector piComposition indicates transformation matrix, b=[b1,b2,b3,b4…bm] indicate the weights of corresponding feature vector.
S24 centered on the coordinate of this feature point, takes its normal direction first to each characteristic point of each face-image
Each n in both sidespThe gray value of a pixel composition of vector in order, then j-th of characteristic point in sample set on i-th of face-image
Local gray level vectorSecondly shade of gray vector is calculatedThen to g 'ijPlace is normalized
Reason obtains normalized gradient vectorThen calculate average gray vectorWith covariance matrix Sj:
It is considered that this standardized grayscale derivative vector meets Gaussian Profile on the whole, then a certain candidate point is corresponded to
Standardized grayscale derivative vector G 'jIt is expressed as with the mahalanobis distance of Average normalized gray scale derivative vectorThis distance reflects the probability that the candidate point comes from this characteristic point,
The corresponding probability in the search process of characteristic point it may determine that difference determines optimal candidate point.
S25 treats positioning face-image first with adaboost algorithms and carries out spy in the search phase of face key point
Levy the initial alignment of point.Adaboost algorithms are the multistage classifier that Face datection is carried out using strong and weak classifiers combination, effect
Well, it can be used for being accurately positioned and arrive face location.After this step determines the initial position of face, subsequent step is then used
In the exact position for finding face, being adjusted to average face model keeps it closest with real target shape, you can looks for
To the optimum position of face key feature points.
S26, the face initial position determined according to S25 steps establish local texture model, are searched carrying out characteristic point position
The position of each key feature points is updated when rope as search criteria.
The renewal process of S27, characteristic point are as follows:Centered on each characteristic point of initial alignment, its normal direction two is taken
Each m in side (m>K) pixel, m indicate that pixel number to be investigated, k indicate the pixel both sides investigated for calculating gradient
Number of pixels.Then the normalized gradient vector that its dimension is 2k is calculated to 2 (m-k)+1 pixel to be investigated.Assuming that
Gj(d) (1≤j≤n) is the normalized gradient vector of d-th pixel in j-th of characteristic point normal direction, then the pixel with
The mahalanobis distance of the local texture model of this feature point isIt will most
New position of the corresponding pixel of small Dist as current signature point, the location variation of all characteristic points are denoted as dx, i.e. dx generations
The relative variation of table new characteristic point position and last time characteristic point position.
S28, old faceform during each characteristic point is updatedIt is aligned with X+dx, obtains changing for attitude parameter s, θ, t
Variable ds,dθ,dt, wherein s indicates that zoom scale, θ indicate that rotation angle, t indicate translation vector, then following formula is utilized to calculate position
Set adjusting parameter dx: Refer to statistical shape model.ThenIts
In,
Find out dxAfterwards, a d is foundbFollowing formula is set to set up:Then shape
The knots modification d of parameter bbFor:db=PTdx。
It calculates knots modification and carries out parameter update later:S is updated to s (1+ds), θ is updated to θ (1+dθ), b is updated to b (1+
db), t is updated to t (1+dt), and byNew shape is calculated, the above step is then repeated
Suddenly until change in shape within a predetermined range, using corresponding feature point coordinates at this time as characteristic point in present frame face-image
Positioning.
After successfully navigating to face key feature points, so that it may to be directed to human eye using the crucial point coordinates got
The judgement of stretching degree is carried out with mouth bar region.
The calculating of human eye stretching degree is salient point (the ordinate α by positioning human eye upper edgeeye) and lower edge is recessed
Point (ordinate βeye), the difference in height of the two is then calculated, by this difference in height divided by the height ω of face area, as eyes
The ratio of shared faceThe ratio can be used as a kind of representation that human eye opens amplitude degree.
Similarly, the difference in height α in mouth bar region can be calculatedmouth-βmouthRelative to the ratio shared by face height ω
ExampleThe ratio can open a kind of representation of amplitude as mouth bar.
As human eye ratio ηeyeWhen less than a scheduled threshold value, then it is assumed that the action of eye closing eyeball occurs in driver.When mouth bar
Ratio ηmouthWhen higher than a predetermined threshold, then it is assumed that the action yawned occurs in driver.If there is Q frame images in P frame images
In the presence of closing one's eyes and action of yawning, then it is assumed that fatigue state occurs in driver.
The above is only some embodiments of the present invention, it is noted that for the ordinary skill people of the art
For member, various improvements and modifications may be made without departing from the principle of the present invention, these improvements and modifications are also answered
It is considered as protection scope of the present invention.
Claims (9)
1. a kind of tired driver based on key point extraction drives monitoring system, which is characterized in that including:
Image capture module is set as acquiring the face-image of driver in real time;
Positioning feature point module is set as positioning the key feature points in every frame face-image;And
State determination module, be set as according to the key feature points positioned in each frame face-image judge current driver's eyes and
The corresponding opening amplitude of face, if there is the opening of Q frame images corresponding driver's eyes and face in continuous P frame face-image
Amplitude is less than corresponding predetermined threshold, then judges that tired driver drives, wherein P and Q is preset value.
2. the tired driver according to claim 1 based on key point extraction drives monitoring system, which is characterized in that described
Image capture module includes camera, infrared light compensating lamp and Video Decoder.
3. the tired driver according to claim 1 based on key point extraction drives monitoring system, which is characterized in that described
Positioning feature point module is using the positioning of ASM algorithms per the key feature points in frame face-image.
4. the tired driver according to claim 1 based on key point extraction drives monitoring system, which is characterized in that the prison
Examining system further includes an image transmission module, when state determination module judgement tired driver drives, selects a frame driver
The face-image that the opening amplitude of eyes and face is less than corresponding predetermined threshold is sent to by described image transmission module on one
Position machine.
5. the tired driver according to claim 1 based on key point extraction drives monitoring system, which is characterized in that the prison
Examining system further includes an alarm module, and when state determination module judgement tired driver drives, the alarm module is sent out
One fatigue warning signal.
6. a kind of tired driver based on key point extraction drives monitoring method, which is characterized in that include the following steps:
S1 acquires the face-image of driver;
S2 is positioned per the key feature points in frame face-image;
S3 judges current driver's eyes and the corresponding opening frame of face according to the key feature points positioned in each frame face-image
Degree, if there have the opening amplitude of the corresponding driver's eyes of Q frame images and face to be less than in continuous P frame face-image to be corresponding pre-
Determine threshold value, then judge that tired driver drives, wherein P and Q is preset value.
7. the tired driver according to claim 6 based on key point extraction drives monitoring method, which is characterized in that described
Step S2 adopts the positioning of ASM algorithms per the key feature points in frame face-image.
8. the tired driver according to claim 6 based on key point extraction drives monitoring method, which is characterized in that described
Step S3 further includes:When judging that tired driver drives, the opening amplitude of a frame driver eyes and face is selected to be less than corresponding pre-
The face-image for determining threshold value is sent to a host computer.
9. the tired driver according to claim 6 based on key point extraction drives monitoring method, which is characterized in that described
Step S3 further includes:When state determination module judgement tired driver drives, a fatigue warning signal is sent out.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810095177.2A CN108363968A (en) | 2018-01-31 | 2018-01-31 | A kind of tired driver driving monitoring system and method based on key point extraction |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201810095177.2A CN108363968A (en) | 2018-01-31 | 2018-01-31 | A kind of tired driver driving monitoring system and method based on key point extraction |
Publications (1)
Publication Number | Publication Date |
---|---|
CN108363968A true CN108363968A (en) | 2018-08-03 |
Family
ID=63007486
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201810095177.2A Pending CN108363968A (en) | 2018-01-31 | 2018-01-31 | A kind of tired driver driving monitoring system and method based on key point extraction |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN108363968A (en) |
Cited By (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109165630A (en) * | 2018-09-19 | 2019-01-08 | 南京邮电大学 | A kind of fatigue monitoring method based on two-dimentional eye recognition |
CN109740477A (en) * | 2018-12-26 | 2019-05-10 | 联创汽车电子有限公司 | Study in Driver Fatigue State Surveillance System and its fatigue detection method |
CN109858553A (en) * | 2019-01-31 | 2019-06-07 | 深圳市赛梅斯凯科技有限公司 | Monitoring model update method, updating device and the storage medium of driving condition |
CN113642426A (en) * | 2021-07-29 | 2021-11-12 | 深圳市比一比网络科技有限公司 | Fatigue detection method and system based on target and key points |
Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN202130312U (en) * | 2011-07-29 | 2012-02-01 | 四川中唯交通科技有限公司 | Driver fatigue driving monitoring device |
CN104361716A (en) * | 2014-10-31 | 2015-02-18 | 新疆宏开电子***集成有限公司 | Method for detecting and reminding fatigue in real time |
CN104408878A (en) * | 2014-11-05 | 2015-03-11 | 唐郁文 | Vehicle fleet fatigue driving early warning monitoring system and method |
CN104688251A (en) * | 2015-03-02 | 2015-06-10 | 西安邦威电子科技有限公司 | Method for detecting fatigue driving and driving in abnormal posture under multiple postures |
CN104951743A (en) * | 2015-03-04 | 2015-09-30 | 苏州大学 | Active-shape-model-algorithm-based method for analyzing face expression |
CN106295600A (en) * | 2016-08-18 | 2017-01-04 | 宁波傲视智绘光电科技有限公司 | Driver status real-time detection method and device |
CN106372621A (en) * | 2016-09-30 | 2017-02-01 | 防城港市港口区高创信息技术有限公司 | Face recognition-based fatigue driving detection method |
CN107229922A (en) * | 2017-06-12 | 2017-10-03 | 西南科技大学 | A kind of fatigue driving monitoring method and device |
-
2018
- 2018-01-31 CN CN201810095177.2A patent/CN108363968A/en active Pending
Patent Citations (8)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN202130312U (en) * | 2011-07-29 | 2012-02-01 | 四川中唯交通科技有限公司 | Driver fatigue driving monitoring device |
CN104361716A (en) * | 2014-10-31 | 2015-02-18 | 新疆宏开电子***集成有限公司 | Method for detecting and reminding fatigue in real time |
CN104408878A (en) * | 2014-11-05 | 2015-03-11 | 唐郁文 | Vehicle fleet fatigue driving early warning monitoring system and method |
CN104688251A (en) * | 2015-03-02 | 2015-06-10 | 西安邦威电子科技有限公司 | Method for detecting fatigue driving and driving in abnormal posture under multiple postures |
CN104951743A (en) * | 2015-03-04 | 2015-09-30 | 苏州大学 | Active-shape-model-algorithm-based method for analyzing face expression |
CN106295600A (en) * | 2016-08-18 | 2017-01-04 | 宁波傲视智绘光电科技有限公司 | Driver status real-time detection method and device |
CN106372621A (en) * | 2016-09-30 | 2017-02-01 | 防城港市港口区高创信息技术有限公司 | Face recognition-based fatigue driving detection method |
CN107229922A (en) * | 2017-06-12 | 2017-10-03 | 西南科技大学 | A kind of fatigue driving monitoring method and device |
Non-Patent Citations (2)
Title |
---|
白中浩等: "基于ASM的多特征融合驾驶员疲劳检测方法", 《电子测量与仪器学报》 * |
韩玉峰等: "一种基于改进的ASM的人脸特征点定位方法", 《计算机科学》 * |
Cited By (6)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN109165630A (en) * | 2018-09-19 | 2019-01-08 | 南京邮电大学 | A kind of fatigue monitoring method based on two-dimentional eye recognition |
CN109740477A (en) * | 2018-12-26 | 2019-05-10 | 联创汽车电子有限公司 | Study in Driver Fatigue State Surveillance System and its fatigue detection method |
CN109740477B (en) * | 2018-12-26 | 2022-12-06 | 联创汽车电子有限公司 | Driver fatigue detection system and fatigue detection method thereof |
CN109858553A (en) * | 2019-01-31 | 2019-06-07 | 深圳市赛梅斯凯科技有限公司 | Monitoring model update method, updating device and the storage medium of driving condition |
CN109858553B (en) * | 2019-01-31 | 2023-12-12 | 锦图计算技术(深圳)有限公司 | Method, device and storage medium for updating driving state monitoring model |
CN113642426A (en) * | 2021-07-29 | 2021-11-12 | 深圳市比一比网络科技有限公司 | Fatigue detection method and system based on target and key points |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN104616438B (en) | A kind of motion detection method of yawning for fatigue driving detection | |
CN108363968A (en) | A kind of tired driver driving monitoring system and method based on key point extraction | |
CN100462047C (en) | Safe driving auxiliary device based on omnidirectional computer vision | |
CN101950355B (en) | Method for detecting fatigue state of driver based on digital video | |
CN202257856U (en) | Driver fatigue-driving monitoring device | |
WO2019145578A1 (en) | Neural network image processing apparatus | |
CN111767900B (en) | Face living body detection method, device, computer equipment and storage medium | |
US20100316265A1 (en) | Face authentication device | |
CN202130312U (en) | Driver fatigue driving monitoring device | |
US9646215B2 (en) | Eye part detection apparatus | |
CN111062292B (en) | Fatigue driving detection device and method | |
CN102254151A (en) | Driver fatigue detection method based on face video analysis | |
CN108734086A (en) | The frequency of wink and gaze estimation method of network are generated based on ocular | |
WO2018218839A1 (en) | Living body recognition method and system | |
CN104331160A (en) | Lip state recognition-based intelligent wheelchair human-computer interaction system and method | |
US20220318369A1 (en) | Face recognition system and method capable of updating registered face templates | |
CN110435672A (en) | One kind is based on scene anticipation safe driving householder method and its system | |
CN109344909A (en) | A kind of personal identification method based on multichannel convolutive neural network | |
JP2005309765A (en) | Image recognition device, image extraction device, image extraction method and program | |
CN114333011B (en) | Network training method, face recognition method, electronic device and storage medium | |
Assiri et al. | Face emotion recognition based on infrared thermal imagery by applying machine learning and parallelism | |
CN113920591A (en) | Medium-distance and long-distance identity authentication method and device based on multi-mode biological feature recognition | |
WO2015037973A1 (en) | A face identification method | |
CN110688872A (en) | Lip-based person identification method, device, program, medium, and electronic apparatus | |
Chen et al. | Video-based face recognition technology for automotive security |
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 | ||
WD01 | Invention patent application deemed withdrawn after publication |
Application publication date: 20180803 |
|
WD01 | Invention patent application deemed withdrawn after publication |