CN101984453B - Human eye recognition system and method - Google Patents
Human eye recognition system and method Download PDFInfo
- Publication number
- CN101984453B CN101984453B CN201010531501.4A CN201010531501A CN101984453B CN 101984453 B CN101984453 B CN 101984453B CN 201010531501 A CN201010531501 A CN 201010531501A CN 101984453 B CN101984453 B CN 101984453B
- Authority
- CN
- China
- Prior art keywords
- eye
- human eye
- human
- left eyes
- target area
- 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.)
- Expired - Fee Related
Links
- 238000000034 method Methods 0.000 title claims abstract description 30
- 238000001514 detection method Methods 0.000 claims abstract description 24
- 210000001747 pupil Anatomy 0.000 claims abstract description 12
- 238000000605 extraction Methods 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 abstract description 3
- 238000010586 diagram Methods 0.000 description 6
- 238000012549 training Methods 0.000 description 3
- IQVNEKKDSLOHHK-FNCQTZNRSA-N (E,E)-hydramethylnon Chemical compound N1CC(C)(C)CNC1=NN=C(/C=C/C=1C=CC(=CC=1)C(F)(F)F)\C=C\C1=CC=C(C(F)(F)F)C=C1 IQVNEKKDSLOHHK-FNCQTZNRSA-N 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 230000004886 head movement Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012360 testing method Methods 0.000 description 2
- BIIBYWQGRFWQKM-JVVROLKMSA-N (2S)-N-[4-(cyclopropylamino)-3,4-dioxo-1-[(3S)-2-oxopyrrolidin-3-yl]butan-2-yl]-2-[[(E)-3-(2,4-dichlorophenyl)prop-2-enoyl]amino]-4,4-dimethylpentanamide Chemical compound CC(C)(C)C[C@@H](C(NC(C[C@H](CCN1)C1=O)C(C(NC1CC1)=O)=O)=O)NC(/C=C/C(C=CC(Cl)=C1)=C1Cl)=O BIIBYWQGRFWQKM-JVVROLKMSA-N 0.000 description 1
- 238000013459 approach Methods 0.000 description 1
- 238000012937 correction Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 210000004709 eyebrow Anatomy 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000005259 measurement Methods 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000002203 pretreatment Methods 0.000 description 1
- 238000012545 processing Methods 0.000 description 1
Images
Landscapes
- Image Analysis (AREA)
Abstract
本发明公开了一种人眼识别系统及方法，所述系统包括：人眼待测区域识别模块，用于从人脸区域上划取左眼和右眼待检测区域；人眼检测模块，用于从左眼和右眼待检测区域检测出左眼和右眼目标区域；人眼权值识别模块，用于通过对左眼和右眼目标区域进行权值计算得到人眼区域，所述权值包括左右眼的大小接近程度、左右眼区域边长与左右眼瞳孔间距比的合理程度和左右眼瞳孔间距连线与水平方向夹角。通过采用本发明所述的人眼识别系统，在检测出左眼和右眼目标区域后，通过三个权值来精确定位人眼，有效解决多个人眼目标区域的定位问题，提高了人眼识别的精确性。
The invention discloses a human eye recognition system and method. The system includes: a human eye detection area identification module, which is used to draw the left eye and right eye detection area from the human face area; the human eye detection module, which uses It is used to detect the left eye and right eye target areas from the left eye and right eye target areas to be detected; the human eye weight recognition module is used to obtain the human eye area by performing weight calculation on the left eye and right eye target areas, and the weight The values include the closeness of the size of the left and right eyes, the reasonableness of the ratio of the side length of the left and right eye area to the distance between the pupils of the left and right eyes, and the angle between the line connecting the distance between the pupils of the left and right eyes and the horizontal direction. By adopting the human eye recognition system described in the present invention, after detecting the target areas of the left eye and the right eye, the human eyes are accurately positioned through three weights, which effectively solves the positioning problem of multiple human eye target areas and improves the accuracy of human eyes. recognition accuracy.
Description
Technical field
The present invention relates generally to the living things feature recognition field, specifically, relate to a kind of auxiliary human eye detection system and method based on eyes structure weights.
Background technology
Development along with computer vision technique, recognition of face and Expression Recognition more and more come into one's own, and human eye detection is as the priori conditions of the pre-treatment step of recognition of face and Expression Recognition, its accuracy of detection and speed directly have influence on precision and the speed of identification, and the accuracy of human eye detection can improve recognition of face and Expression Recognition accuracy and automaticity effectively.At present, most Expression Recognition and face recognition algorithms are all the human eyes of first having good positioning, and then carry out the calculating of normalization, feature calculation and unique point according to the positional information of human eye, and then carry out Expression Recognition or recognition of face.
With reference to figure 1, Fig. 1 shows the method flow diagram of at present main eye recognition, and its key step comprises: gather the picture pedestrian's face of going forward side by side and detect, obtain people's face rectangular area; With the rectangular area, draw and to get left eye and right eye zone to be detected; In extracted region Haar feature to be detected (simple rectangular characteristic), and collect non-human eye area and inhuman eye pattern sheet, extract the Haar feature; According to the Haar feature of human eye picture and inhuman eye pattern sheet, training AdaBoost human eye detection model; The test picture obtains left eye and right eye zone to be detected, calculates the Haar feature, input AdaBoost human eye detection model; If the zone, left and right detects respectively a human eye, algorithm success, otherwise failure.Existing human eye location algorithm, although obtained certain effect, but still have problems, such as locating out of true, can't locating a plurality of human eyes, so many times still need a large amount of manual correction in the human eye position fixing process, this problem causes being difficult to realize the robotization of Expression Recognition and recognition of face.
Therefore, be necessary to propose a kind of recognition system and the method that can accurately locate human eye.
Summary of the invention
In order to address the above problem, the invention provides a kind of human eye recognition system, described system comprises: human eye regional identification module to be measured is used for drawing from human face region and gets left eye and right eye zone to be detected; The human eye detection module is used for detecting from left eye and right eye zone to be detected left eye and right eye target area; Human eye weights identification module, be used for calculating human eye area by weights are carried out in left eye and right eye target area, described weights comprise resonable degree and right and left eyes interocular distance line and the horizontal direction angle of big or small degree of closeness, the right and left eyes zone length of side and the right and left eyes interocular distance ratio of right and left eyes.
The present invention also provides a kind of eye recognition method, and described method comprises: obtain human face region, draw from described human face region and get left eye and right eye zone to be detected; Detect left eye and right eye target area from left eye and right eye zone to be detected; Weights are carried out in left eye and right eye target area calculate human eye area, described weights comprise resonable degree and right and left eyes interocular distance line and the horizontal direction angle of big or small degree of closeness, the right and left eyes zone length of side and the right and left eyes interocular distance ratio of right and left eyes.
By adopting human eye recognition system of the present invention, after detecting left eye and right eye target area, the resonable degree of the big or small degree of closeness by right and left eyes, the right and left eyes zone length of side and right and left eyes interocular distance ratio and right and left eyes interocular distance line and these three weights of horizontal direction angle are located human eye, effectively solve the orientation problem of a plurality of human eyes target area, improved accuracy and the automaticity of eye recognition.
Description of drawings
Fig. 1 shows the process flow diagram of human eye recognition system;
Fig. 2 shows the schematic diagram of human eye detection Haar feature;
Fig. 3 shows human eye recognition system structural representation and the process flow diagram according to the embodiment of the present invention;
Fig. 4, Fig. 5 show the human eye sample area schematic diagram according to the embodiment of the present invention.
Embodiment
The below describes embodiments of the invention in detail, and the example of described embodiment is shown in the drawings, is exemplary below by the embodiment that is described with reference to the drawings, and only is used for explaining the present invention, and can not be interpreted as limitation of the present invention.
With reference to figure 3, the human eye recognition system of the embodiment of the present invention comprises: human eye regional identification module 100 to be measured, human eye detection module 200 and human eye weights identification module 300.
Described human eye regional identification module 100 to be measured is used for drawing from human face region and gets left eye and right eye zone to be detected.
Described human eye detection module 200 is used for detecting from left eye and right eye zone to be detected left eye and right eye target area.In embodiments of the present invention, described human eye detection module comprises human eye feature extraction module and eye recognition module, be used for to extract the feature in left eye and right eye zone to be detected by described human eye feature extraction module, the non-human eye area of feature that is used for getting rid of described extraction by described eye recognition module is to obtain left eye and right eye target area.
Described human eye weights identification module 300, be used for calculating human eye area by weights are carried out in left eye and right eye target area, described weights comprise resonable degree and right and left eyes interocular distance line and the horizontal direction angle of big or small degree of closeness, the right and left eyes zone length of side and the right and left eyes interocular distance ratio of right and left eyes.
According to process flow diagram shown in Figure 3, can realize by following step and method the eye recognition of described human eye recognition system.
At step S01, by detecting at the enterprising pedestrian's face of the portrait picture that gathers, obtain people's face rectangular area.
At step S02, draw from human face region and get left eye and right eye zone to be detected.
At step S03, extract the feature in left eye and right eye zone to be detected.In embodiments of the present invention, can be by extracting the Haar feature as the feature of identification people's face and human eye, the Haar feature is divided three classes usually: edge feature, linear feature, central feature and diagonal line feature, be combined into feature templates, as shown in Figure 2, adularescent and two kinds of rectangles of black in feature templates, and can be by the eigenwert that defines this template the white rectangle pixel and deduct the black rectangle pixel and, therefore can distinguish human eye and non-human eye by select suitable feature on the Haar of picture characteristic set.
At step S04, get rid of non-human eye area in the feature of described extraction, to obtain left eye and right eye target area.In embodiments of the present invention, can pass through the AdaBoost algorithm, adopt wipe-out mode to remove non-human eye area based on the Haar characteristic information, obtain left eye and right eye target area.The AdaBoost algorithm is a kind of method of cascade sort, and it is combined some more weak sorting techniques, the very strong sorting technique that combination makes new advances.In embodiments of the present invention, at first, can select effective Haar feature by the AdaBoost algorithm from the Haar feature and train the right and left eyes detecting device.Training sample is divided into positive example sample and negative data, wherein the positive example sample refers to target sample to be checked (left eye and right eye), negative data refers to other any image, all samples pictures all are normalized to same size, size in samples pictures described in one embodiment of the invention can be 20*20, the size of described samples pictures can be come to determine as required, can also be 30*30 or other suitable sizes.After sorter has been trained, can be applied to the detection in the zone to be detected of left eye in input picture and right eye, described zone to be detected has identical size with training sample.In the image detection process, with image convergent-divergent, the then subwindow of exhaustive each 20*20 of the picture after convergent-divergent successively in proportion.By each grade sorter, in which floor detection, most candidate region just has been excluded window so in front successively, and all the zone by each grade detection of classifier is the target area.Then, adopt wipe-out mode to remove non-human eye area based on the Haar characteristic information, what testing process finished to obtain is all zones that are not excluded.But in the bad situation of light, the zone such as canthus eyebrow and human eye have similar Haar feature, are easy to be identified as human eye by mistake, the situation that people's face one side identifies a plurality of human eyes occurs.
At step S05, weights are carried out in left eye and right eye target area calculate human eye area, the present invention calculates accurate human eye area by three weights, described three weights comprise big or small degree of closeness, the right and left eyes zone length of side of right and left eyes and resonable degree and right and left eyes interocular distance line and the horizontal direction angle of right and left eyes interocular distance ratio, by judge two human eye degrees of closeness, eyes size and interpupillary distance ratio whether near 0.6 and interpupillary line and horizontal sextant angle obtain a human eye area the most accurate near these three weights of level.In the embodiment of the present invention, described weight calculation method is:
The big or small degree of closeness of described left eye and right eye: W
1=1-|S
l-S
r|/(S
l+ S
r), S
lRepresent the left eye region length of side, S
rRepresent the right eye region length of side.If S
lAnd S
rSize approaches so | S
l-S
r| near 0, thus W
1Close to 1, therefore, for the big or small degree of closeness W of described left eye and right eye
1Value more close to 1, the size of left eye and right eye is more approaching.
The length of side in described right and left eyes zone and the resonable degree of right and left eyes interocular distance ratio: W
2=1-|L
p-L|/(L
p+ L), L
pRepresent the spacing of two pupils, L is by L=(S
l+ S
r) * 5/6 calculates and get, in embodiments of the present invention, than whether judges described weights W near 0.6 by the regional length of side of right and left eyes with the right and left eyes interocular distance
2Resonable degree, L wherein
pRepresent the distance of two pupil actual measurements, L is the reference value that calculates by the empirical value that the right and left eyes length of side is interocular distance 0.6, with reference to figure 4, L
pMore approaching with L, W
2Value more close to 1, the length of side in described right and left eyes zone than more reasonable, thinks that right and left eyes more meets normal eye's standard with the right and left eyes interocular distance.
The angle of described right and left eyes pupil center's line and horizontal direction: W
3=D
x/ (D
x+ D
y), D
xRepresent the horizontal range of two pupils; D
yRepresent the vertical range of two pupils, with reference to figure 5, if people's eye coordinates (x of pair of right and left human eye
l, y
l) and (x
r, y
r) the difference in height of two eyes be D
x, D
yIf human eye is in normal condition, D
yClose to 0, described weights W
3Close to 1, described weights W
3Be 1 to the maximum.
For above each weights, too small irrational data appear sometimes, can as required it be screened out, for example, keep the data near 1, give less negative value with too small data, the data of-10 left and right in the end during Weight selected, will not considered in for example-10.
For better explanation the present invention, below will be elaborated with the embodiment based on the frightened Expression Recognition of head movement.Described embodiment need to set up an expression database in advance, and in database, each expression comprises start frame and exaggerates frame most, and the equal mark of the expression in database.
In identifying, at first, take out all expression start frames and exaggerate frame most, advanced pedestrian's face detects in every width picture, the rectangle human face region that obtains detecting, and the height of rectangular area: height, wide: width.
Then, determine the human eye target area.Get top left region: (x ∈ [0, width/2], y ∈ [0,0.6*height]) from the human eye detection algorithm of top left region application based on Haar feature and AdaBoost sorter, detects M human eye; In like manner, use identical algorithms in right regions (x ∈ [width/2, width], y ∈ [0,0.6*height]), N human eye detected.
Then, determine accurate human eye area.If M*N=0; The failure of human eye detection algorithm; If M*N=1; The success of human eye detection algorithm; If M*N＞1, all human eyes that detect form M*N to human eye; According to the every pair of human eye that detects calculate three weights and.Weighting value and maximum human eye are to as net result.The success of human eye detection algorithm.
According to the human eye that has detected, calculate the head movement information characteristics.Adopt sorter that feature is classified, and then whether identification is frightened expression.
Above human eye recognition system of the present invention and method are described in detail, the resonable degree of the big or small degree of closeness by right and left eyes, the right and left eyes zone length of side and right and left eyes interocular distance ratio and right and left eyes interocular distance line and these three weights of horizontal direction angle are located human eye, effectively solve the orientation problem of a plurality of human eyes target area, improved accuracy and the automaticity of eye recognition.
One of ordinary skill in the art will appreciate that and realize that all or part of step that above-described embodiment method is carried is to come the relevant hardware of instruction to complete by program, described program can be stored in a kind of computer-readable recording medium, this program comprises step of embodiment of the method one or a combination set of when carrying out.
In addition, each functional unit in each embodiment of the present invention can be integrated in a processing module, can be also that the independent physics of unit exists, and also can be integrated in a module two or more unit.Above-mentioned integrated module both can adopt the form of hardware to realize, also can adopt the form of software function module to realize.If described integrated module realizes with the form of software function module and during as independently production marketing or use, also can be stored in a computer read/write memory medium.
The above-mentioned storage medium of mentioning can be ROM (read-only memory), disk or CD etc.
The above is only embodiments of the present invention; should be pointed out that for those skilled in the art, under the prerequisite that does not break away from the principle of the invention; can also make some improvements and modifications, these improvements and modifications also should be considered as protection scope of the present invention.
Claims (10)
1. human eye recognition system, described system comprises:
Human eye regional identification module to be measured is used for drawing from human face region and gets left eye and right eye zone to be detected;
The human eye detection module is used for detecting from left eye and right eye zone to be detected left eye and right eye target area, and the quantity of detected left eye target area is M, and the quantity of detected right eye target area is N;
Human eye weights identification module is used for as M*N〉1 the time, left eye target area and right eye target area are made up, obtain M*N to the target human eye area; Weights are all carried out respectively in every pair of human eye target area calculate, described weights comprise resonable degree and right and left eyes interocular distance line and the horizontal direction angle of big or small degree of closeness, the right and left eyes zone length of side and the right and left eyes interocular distance ratio of right and left eyes; Big or small degree of closeness by judging two people's right and left eyes, the right and left eyes zone length of side and right and left eyes interocular distance than whether near 0.6 and right and left eyes interocular distance line and horizontal direction angle whether select a pair of human eye target area as human eye area near level.
2. system according to claim 1, it is characterized in that, described human eye detection module comprises human eye feature extraction module and eye recognition module, described human eye feature extraction module be used for to extract the Haar feature in left eye and right eye zone to be detected, described eye recognition module is used for getting rid of the non-human eye area of Haar feature of described extraction, to obtain left eye and right eye target area.
3. system according to claim 2, is characterized in that, described eye recognition module detects identification object region by the AdaBoost algorithm.
4. system according to claim 1, is characterized in that, the big or small degree of closeness W of described right and left eyes
1=1-|S
l-S
r|/(S
l+ S
r), S
lRepresent the left eye region length of side, S
rRepresent the right eye region length of side; The resonable degree W of the right and left eyes zone length of side and right and left eyes interocular distance ratio
2=1-|L
p-L|/(L
p+ L), L
pRepresent the spacing of two pupils, L=(S
l+ S
r) * 5/6; Right and left eyes interocular distance line and horizontal direction angle, wherein W
3=D
x/ (D
x+ D
y), D
xRepresent the horizontal range of two pupils, D
yRepresent the vertical range of two pupils.
5. system according to claim 4, is characterized in that, by described weights addition is obtained weights and, weights and maximum a pair of human eye target area are defined as human eye area.
6. eye recognition method, described method comprises:
Obtain human face region, draw from described human face region and get left eye and right eye zone to be detected;
Detect left eye and right eye target area from left eye and right eye zone to be detected, the quantity of detected left eye target area is M, and the quantity of detected right eye target area is N;
As M*N〉1 the time, left eye target area and right eye target area are made up, obtain M*N to the target human eye area;
Weights are all carried out respectively in every pair of human eye target area calculate, described weights comprise resonable degree and right and left eyes interocular distance line and the horizontal direction angle of big or small degree of closeness, the right and left eyes zone length of side and the right and left eyes interocular distance ratio of right and left eyes; Big or small degree of closeness by judging two people's right and left eyes, the right and left eyes zone length of side and right and left eyes interocular distance than whether near 0.6 and right and left eyes interocular distance line and horizontal direction angle whether select a pair of human eye target area as human eye area near level.
7. method according to claim 6, is characterized in that, the method that obtains described left eye and right eye target area comprises: the Haar feature of extracting left eye and right eye zone to be detected; Get rid of non-human eye area in the Haar feature of described extraction, to obtain left eye and right eye target area.
8. method according to claim 7, is characterized in that, in the Haar feature of the described extraction of eliminating, the method for non-human eye area is the AdaBoost classification.
9. method according to claim 6, is characterized in that, the big or small degree of closeness W of described right and left eyes
1=1-|S
l-S
r|/(S
l+ S
r), S
lRepresent the left eye region length of side, S
rRepresent the right eye region length of side; The resonable degree W of the right and left eyes zone length of side and right and left eyes interocular distance ratio
2=1-|L
p-L|/(L
p+ L), L
pRepresent the spacing of two pupils, L=(S
l+ S
r) * 5/6; Right and left eyes interocular distance line and horizontal direction angle W
3=D
x/ (D
x+ D
y), D
xRepresent the horizontal range of two pupils, D
yRepresent the vertical range of two pupils.
10. method according to claim 9, is characterized in that, by described weights addition is obtained weights and, weights and maximum a pair of human eye target area are defined as human eye area.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010531501.4A CN101984453B (en) | 2010-11-02 | 2010-11-02 | Human eye recognition system and method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201010531501.4A CN101984453B (en) | 2010-11-02 | 2010-11-02 | Human eye recognition system and method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN101984453A CN101984453A (en) | 2011-03-09 |
| CN101984453B true CN101984453B (en) | 2013-05-08 |
Family
ID=43641622
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201010531501.4A Expired - Fee Related CN101984453B (en) | 2010-11-02 | 2010-11-02 | Human eye recognition system and method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN101984453B (en) |
Families Citing this family (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102426644B (en) * | 2011-08-24 | 2013-07-10 | 湖南领创智能科技有限公司 | Human eye cross localization method based on GentleBoost machine learning |
| CN102830797B (en) * | 2012-07-26 | 2015-11-25 | 深圳先进技术研究院 | A kind of man-machine interaction method based on sight line judgement and system |
| CN102867172B (en) * | 2012-08-27 | 2016-12-21 | Tcl集团股份有限公司 | A kind of human-eye positioning method, system and electronic equipment |
| CN104346621A (en) * | 2013-07-30 | 2015-02-11 | 展讯通信(天津)有限公司 | Method and device for creating eye template as well as method and device for detecting eye state |
| CN104679225B (en) * | 2013-11-28 | 2018-02-02 | 上海斐讯数据通信技术有限公司 | Screen adjustment method, screen adjustment device and the mobile terminal of mobile terminal |
| CN104732202A (en) * | 2015-02-12 | 2015-06-24 | 杭州电子科技大学 | Method for eliminating influence of glasses frame during human eye detection |
| CN104933344B (en) * | 2015-07-06 | 2019-01-04 | 北京中科虹霸科技有限公司 | Mobile terminal user identity authentication device and method based on multi-biological characteristic mode |
| CN107679448B (en) | 2017-08-17 | 2018-09-25 | 平安科技（深圳）有限公司 | Eyeball action-analysing method, device and storage medium |
| WO2019116078A1 (en) * | 2017-12-13 | 2019-06-20 | Medical Diagnostech (Pty) Ltd | System and method for obtaining a pupil response profile |
| CN109118626B (en) * | 2018-08-08 | 2022-09-13 | 腾讯科技（深圳）有限公司 | Lock control method and device, storage medium and electronic device |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1731418A (en) * | 2005-08-19 | 2006-02-08 | 清华大学 | Robust method for precise eye localization in complex background images |
| CN101059836A (en) * | 2007-06-01 | 2007-10-24 | 华南理工大学 | Human eye positioning and human eye state recognition method |
Family Cites Families (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP3425092B2 (en) * | 1998-11-04 | 2003-07-07 | 株式会社エヌ・ティ・ティ・データ | Pattern recognition method |
| JP5014307B2 (en) * | 2008-10-29 | 2012-08-29 | 三菱電機株式会社 | Image recognition system and air conditioner |
-
2010
- 2010-11-02 CN CN201010531501.4A patent/CN101984453B/en not_active Expired - Fee Related
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1731418A (en) * | 2005-08-19 | 2006-02-08 | 清华大学 | Robust method for precise eye localization in complex background images |
| CN101059836A (en) * | 2007-06-01 | 2007-10-24 | 华南理工大学 | Human eye positioning and human eye state recognition method |
Non-Patent Citations (4)
| Title |
|---|
| JP特开2000-137818A 2000.05.16 |
| JP特开2010-108140A 2010.05.13 |
| Shangfei Wang et al.Analysis of Relationships between Color and Emotion by Classification based on Associations.《2008 International Conference on Computer Science and Software Engineering》.2008,269-272. * |
| Shangfei Wang et al.Infrared Facial Expression Recognition Using Wavelet Transform.《2008 International Symposium on Computer Science and Computational Technology》.2008,327-330. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN101984453A (en) | 2011-03-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN101984453B (en) | Human eye recognition system and method | |
| CN108334848B (en) | Tiny face recognition method based on generation countermeasure network | |
| WO2018072233A1 (en) | Method and system for vehicle tag detection and recognition based on selective search algorithm | |
| Yang et al. | Robust superpixel tracking | |
| CN105160317B (en) | A method for gender recognition of pedestrians based on regional blocks | |
| CN105205486B (en) | A kind of automobile logo identification method and device | |
| CN103034852B (en) | The detection method of particular color pedestrian under Still Camera scene | |
| CN103473571B (en) | Human detection method | |
| CN105096307B (en) | The method of detection object in paired stereo-picture | |
| CN104318225B (en) | Detection method of license plate and device | |
| CN101751559B (en) | Method for detecting skin stains on face and identifying face by utilizing skin stains | |
| CN104424466A (en) | Object detection method, object detection device and image pickup device | |
| CN102289660A (en) | Method for detecting illegal driving behavior based on hand gesture tracking | |
| KR101697161B1 (en) | Device and method for tracking pedestrian in thermal image using an online random fern learning | |
| CN105046206B (en) | Based on the pedestrian detection method and device for moving prior information in video | |
| CN106203237A (en) | The recognition methods of container-trailer numbering and device | |
| CN106203401B (en) | A fast detection method of lane lines | |
| CN111027481A (en) | Behavior analysis method and device based on human body key point detection | |
| CN106407958A (en) | Double-layer-cascade-based facial feature detection method | |
| CN106250824A (en) | Vehicle window localization method and system | |
| CN105279492A (en) | Iris identification method and device | |
| CN108388871A (en) | A kind of vehicle checking method returned based on vehicle body | |
| CN103390151A (en) | Face detection method and device | |
| CN114565675A (en) | A method for removing dynamic feature points in the front end of visual SLAM | |
| CN104268595A (en) | General object detecting method and system |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| C14 | Grant of patent or utility model | ||
| GR01 | Patent grant | ||
| CF01 | Termination of patent right due to non-payment of annual fee |
Granted publication date: 20130508 Termination date: 20211102 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |