CN108921010A - A kind of pupil detection method and detection device - Google Patents
A kind of pupil detection method and detection device Download PDFInfo
- Publication number
- CN108921010A CN108921010A CN201810459543.8A CN201810459543A CN108921010A CN 108921010 A CN108921010 A CN 108921010A CN 201810459543 A CN201810459543 A CN 201810459543A CN 108921010 A CN108921010 A CN 108921010A
- Authority
- CN
- China
- Prior art keywords
- pupil
- eyes image
- image
- module
- value
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Granted
Links
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/16—Human faces, e.g. facial parts, sketches or expressions
- G06V40/168—Feature extraction; Face representation
- G06V40/171—Local features and components; Facial parts ; Occluding parts, e.g. glasses; Geometrical relationships
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/136—Segmentation; Edge detection involving thresholding
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T7/00—Image analysis
- G06T7/10—Segmentation; Edge detection
- G06T7/155—Segmentation; Edge detection involving morphological operators
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30196—Human being; Person
- G06T2207/30201—Face
Landscapes
- Engineering & Computer Science (AREA)
- Computer Vision & Pattern Recognition (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Theoretical Computer Science (AREA)
- Health & Medical Sciences (AREA)
- Oral & Maxillofacial Surgery (AREA)
- General Health & Medical Sciences (AREA)
- Human Computer Interaction (AREA)
- Multimedia (AREA)
- Image Analysis (AREA)
- Image Processing (AREA)
Abstract
The present invention relates to technical field of image processing more particularly to a kind of pupil detection method, this method to include：According to facial image, eyes image is obtained；Opening operation processing is carried out to the eyes image of acquisition；The minimum gray value and average gray for calculating the eyes image after opening operation is handled divide pupil using adaptive threshold fuzziness method；According to segmentation result, it is fitted pupil elliptic equation.The invention further relates to a kind of pupil detectors, including：Ocular extraction module, opening operation processing module, adaptive threshold fuzziness module and fitting module.Pupil detection method and detection device provided by the invention can be used for the pupil detection in image, and detection speed is fast, and accuracy rate is high.
Description
Technical field
The present invention relates to technical field of image processing more particularly to a kind of pupil detection methods and detection device.
Background technique
Traditional computer input mode is excessively single, and the mode of operation of keyboard and mouse does not adapt to the need of special population
It asks, human-computer interaction future thrust is more efficient, the natural human-computer interaction sides such as voice, gesture and sight movement
Formula.A kind of mobile scheme as human-computer interaction of sight, principle is to issue instruction using pupil position variation, when human eye is being seen
When in kind, always there is a direction of gaze, this direction can be used as input information and operate to computer.Such interaction side
Formula can liberate the both hands of user, for there is the disabled person of limb defects to be also very helpful.
In addition, in real time, accurate, robust pupil detection and be located in traffic safety, public safety, psychology and criminal investigation
The demand in equal fields is also more and more, and the size of pupil, location status can reflect the part psychological condition and physiological status of people.Than
Such as, it may determine that driver's fatigue degree by detecting the state of pupil, reduce traffic accident；During interrogation,
The pupil status of convict can be used for evaluating the really degree of its confession content.Pupil detection also has in psychiatry research extensively
General application.Therefore, how accurately, robust, in real time detection pupil status just becoming computer vision, pattern-recognition in recent years
One of the synthesis hot issue that equal fields are studied.
Currently, the existing usual calculating process of pupil detection method is complicated, precision is not high or stability is poor, standard can not achieve
True extraction information.
Summary of the invention
Technical purpose of the invention is to provide a kind of accurate, robust, real-time pupil status detection method and detection dress
It sets.
In order to achieve the above technical purposes, the present invention provides a kind of pupil detection methods, including：
S1, according to facial image, obtain eyes image；
S2, opening operation processing is carried out to the eyes image that step S1 is obtained；
S3, the minimum gray value and average gray for calculating the eyes image after opening operation is handled, using adaptive thresholding
It is worth automatic Segmentation pupil；
S4, according to the segmentation result of step S3, be fitted pupil elliptic equation.
Preferably, in the step S1, using adaptive lifting scheme, human eye area is found by classifier, obtains eye
Portion's image.
Preferably, the step S1 further includes carrying out median filter process to obtained eyes image.
Preferably, when dividing pupil in the step S3, eyes image gray value is greater than threshold according to threshold value by threshold value
The part of value is set to maximum value 255, and the part less than or equal to threshold value is set to minimum value 0.
Preferably, in the adaptive threshold fuzziness method of the step S3, threshold value T is calculated using following formula：
T=Min (g (x, y))+Avg (in (x, y))/n；
Wherein, g (x, y) is by opening operation treated eyes image, and in (x, y) is the eye figure obtained in step S1
Picture, Min (g (x, y)) are eyes image minimum gray value, and Avg (in (x, y)) is eyes image average gray, and n is setting
Detection parameters.
Preferably, pupil elliptic equation is fitted using least square method in the step S4, the formula of fitting is：
Wherein, A, B, C, D, E are the polynomial parameters after fitting, (xi, yi) it is the pupil that segmentation pupil obtains in step S3
Zone boundary point, i=1 ..., m, m be pupil region boundary point total number.
The present invention also provides a kind of pupil detectors, including：
Ocular extraction module, for obtaining eyes image according to facial image；
Opening operation processing module, the eyes image for obtaining to ocular extraction module carry out opening operation processing；
Adaptive threshold fuzziness module, the minimum gray value of treated for calculating opening operation processing module eyes image
And average gray, divide pupil using adaptive threshold fuzziness method；
Fitting module is fitted pupil elliptic equation for the segmentation result according to adaptive threshold fuzziness module.
Preferably, the ocular extraction module utilizes adaptive lifting scheme, finds human eye area by classifier,
Obtain eyes image.
Preferably, the adaptive threshold fuzziness module is used for threshold value, and according to threshold value by eyes image gray value
Part greater than threshold value is set to maximum value 255, and the part less than or equal to threshold value is set to minimum value 0；The adaptive threshold fuzziness
Module calculates threshold value T using following formula：
T=Min (g (x, y))+Avg (in (x, y))/n；
Wherein, g (x, y) is by the opening operation processing module treated eyes image, and in (x, y) is the eye
The eyes image that region extraction module obtains, Min (g (x, y)) are eyes image minimum gray value, and Avg (in (x, y)) is eye
Image grayscale average value, n are the detection parameters of setting.
Preferably, the fitting module is fitted pupil elliptic equation using least square method, and the formula of fitting is：
Wherein, A, B, C, D, E are the polynomial parameters after fitting, (xi, yi) it is the adaptive threshold fuzziness module segmentation
The boundary point of pupil region afterwards, i=1 ..., m, m be pupil region boundary point total number.
Above-mentioned technical proposal of the invention has the following advantages that：The present invention provides a kind of pupil detection methods, by face
Eyes image is obtained in image, after the processing of morphological image opening operation, divides pupil using adaptive threshold fuzziness method,
Pupil after fitting segmentation, finds out pupil elliptic equation, pupil detection method calculating process provided by the invention is simple, operation is fast
Degree is fast, accuracy is high, and stability is good.
The present invention also provides a kind of pupil detector, including ocular extraction module, opening operation processing module, from
Threshold segmentation module and fitting module are adapted to, can be used for real-time, accurate, robust detection pupil, obtain the parameter in relation to pupil.
Detailed description of the invention
Fig. 1 is pupil detection method flow chart in the embodiment of the present invention；
Fig. 2 a to Fig. 2 c is detection pupil result figure in the embodiment of the present invention；
Fig. 3 is pupil detector schematic diagram in the embodiment of the present invention.
In figure：100：Ocular extraction module；200：Opening operation processing module；300：Adaptive threshold fuzziness module；
400：Fitting module.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiments of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, ordinary skill people
Member's every other embodiment obtained without making creative work, shall fall within the protection scope of the present invention.
As shown in Figure 1, a kind of pupil detection method provided in an embodiment of the present invention, specifically includes following steps：
S1, according to facial image, obtain eyes image.
In order to improve the precision of pupil extraction, it is close red for being used to shoot the image capture device of facial image in the present embodiment
Outer video camera, the facial image size of shooting are 1280*720.Obviously, in other embodiments, other images can also be passed through
It acquires equipment and obtains facial image, the size of image can also be adjusted according to actual needs.
During human eye detection, the selection of characteristics of image is particularly significant, and characteristics of image will calculate quick, robustness
By force, the robustness and real-time of detection can be just taken into account in this way, selected spy of Ha Er (Haar) feature as image in the present invention
Sign, and Lis Hartel sign can assist calculating with integral image.
Preferably, it is found using adaptive boosting (AdaBoost) method by classifier to accelerate image processing speed
Human eye area, obtains eyes image, which can be trained in advance using method known to a person skilled in the art.
Region of the eyes image as pupil detection is chosen, the time of pupil detection can be effectively saved.Especially for
In the treatment process of consecutive image, be remarkably improved image processing speed, if the pupil region detected does not disappear, persistently with
The variation of the track pupil region；If the pupil region detected disappears, then carries out full figure retrieval, eyes image is obtained.
It is further preferred that step S1 further includes carrying out median filter process to obtained eyes image.Median filtering is
A kind of nonlinear smoothing technology, it sets the gray value of each pixel to all pixels point gray scale in the vertex neighborhood window
The intermediate value of value, the true value for making the pixel value of surrounding close, to eliminate isolated noise spot.The size of neighborhood window can basis
Actual needs selection.
S2, opening operation processing is carried out to the eyes image that step S1 is obtained.
Morphological image, which refers to, to be extracted with mathematical Morphology Algorithm to the picture content expressed and description image is useful, such as
Boundary, bone etc., morphological image can also be filtered image, trim.Due to existing in eyes image at pupil
There is the interference of Purkinje image, can be handled with the method for morphological image.
Morphological image operation includes burn into expansion, opening operation and closed operation.Wherein, corrosion refers to a kind of elimination side
Boundary's point, the process for shrinking boundary internally can be used to eliminate small and meaningless object.Expansion refers to connect with object
All background dots of touching are merged into the object, make boundary to the process of outside expansion, can be used to fill up the cavity in object.
Opening operation refers to first corroding the operation expanded afterwards, can eliminate wisp, separating objects or smooth larger object at very thin point
Boundary, unobvious its area of change.
Opening operation processing in step S2 includes first corroding reflation to eyes image, opens fortune by first corrode reflation
Calculation handles the interference that can eliminate Purkinje image in pupil image.
S3, the minimum gray value and average gray for calculating step S2 treated eyes image, using adaptive threshold
Automatic Segmentation pupil.
Preferably, when dividing pupil in the present embodiment, first threshold value, then according to determining threshold value, by eyes image
The part that gray value is greater than threshold value is set to maximum value 255, and the part less than or equal to threshold value is set to minimum value 0, after being divided
Pupil region.
It is further preferred that the adaptive threshold fuzziness method of step S3 calculates threshold value T using following formula：
T=Min (g (x, y))+Avg (in (x, y))/n；
Wherein, g (x, y) is by step S2 opening operation treated eyes image, and in (x, y) is original image picture value, that is, is walked
The eyes image obtained in rapid S1, Min (g (x, y)) are eyes image minimum gray value, and Avg (in (x, y)) is eyes image ash
Average value is spent, n is the detection parameters of setting, related to image capture device.
In some embodiments of the invention, the detection parameters can be preset according to image capture device, such as close
Thermal camera, the common value range of n are 6-9.
In other embodiments of the invention, the detection of image capture device can also be joined in advance in the following way
Number n is measured：
Using image capture device acquire face and pupil picture several；
Different detection parameters n is adjusted, pupil detection is carried out using different n, if the corresponding detection probability of n is greater than
99%, then n is determined.
Since the threshold value in adaptive threshold fuzziness method is obtained according to image itself, the corresponding threshold value of each pixel may
Not identical, avoiding the factors such as illumination unevenness, burst noise leads to the division of target area and background area mistake.
S4, according to the segmentation result of step S3, be fitted pupil elliptic equation.
Preferably, pupil elliptic equation is fitted using least square method in S4, the formula of fitting is：
Wherein, A, B, C, D, E are the polynomial parameters after fitting, (xi, yi) it is to divide the pupil obtained after pupil in step S3
Bore region boundary point, i=1 ..., m, m be pupil region boundary point total number.
After fitting obtains pupil elliptic equation, the coordinate of pupil and the face of pupil can be solved by pupil elliptic equation
Product, obtains the position of pupil and the size of pupil.As shown in Fig. 2 a to Fig. 2 c, Fig. 2 a to Fig. 2 c is pupil detection in the present embodiment
Method detects the result figure of pupil, and the white circle in figure indicates the pupil elliptic equation curve that fitting obtains, internal to surround
Area be pupil size, the position at pupil center is then indicated at white crosses center.
It should be noted that the software code of the pupil detection method in the present embodiment is all programmed reality with VC++
It is existing, it can also be realized in other embodiments using other modes certainly.
To sum up, a kind of pupil detection method based on least square method and compression tracking is present embodiments provided, can be used for
Pupil detection in the facial image of near-infrared video camera shooting, since image resolution ratio is higher, so being handled using echelonization
Method, first look for human eye area, after by morphological image process, eliminate the interference of Purkinje image in pupil image, then benefit
Divide pupil with adaptive threshold fuzziness method, last least square method fits pupil elliptic equation, according to pupil ellipse side
Journey can find out the area and coordinate of pupil, obtain the size and location of pupil, and the pupil detection method is in 720P resolution ratio
Under the conditions of can achieve the detection speed of 25fps, error is below 0.5 pixel.
As shown in figure 3, the present embodiment additionally provides a kind of pupil detector, including ocular extraction module 100, open
Calculation process module 200, adaptive threshold fuzziness module 300 and fitting module 400.
Ocular extraction module 100 is used to obtain eyes image according to facial image.
Preferably, ocular extraction module 100 chooses Lis Hartel and levies feature as facial image, using adaptively mentioning
Lifting method finds human eye area by classifier, obtains eyes image.
It is further preferred that ocular extraction module 100 further includes carrying out at median filtering to obtained eyes image
Reason, eliminates isolated noise spot.
The eyes image that opening operation processing module 200 is used to obtain ocular extraction module 100 carries out at opening operation
Reason.Opening operation processing includes first corroding reflation to eyes image.
Adaptive threshold fuzziness module 300 is used to calculate opening operation processing module 200 gray scale of treated eyes image
Minimum value and average gray divide pupil using adaptive threshold fuzziness method.
Preferably, adaptive threshold fuzziness module 300 is used for threshold value, and according to threshold value that eyes image gray value is big
It is set to maximum value 255 in the part of threshold value, the part less than or equal to threshold value is set to minimum value 0.
It is further preferred that adaptive threshold fuzziness module 300 calculates threshold value T using following formula：
T=Min (g (x, y))+Avg (in (x, y))/n；
Wherein, g (x, y) is by treated the eyes image of opening operation processing module 200, and in (x, y) is original image picture value,
I.e. ocular extraction module 100 obtain eyes image, Min (g (x, y)) be eyes image minimum gray value, Avg (in (x,
It y)) is eyes image average gray, n is the detection parameters of setting, related to the acquisition image capture module of face.
Fitting module 400 is used for the segmentation result according to adaptive threshold fuzziness module 300, is fitted pupil elliptic equation.
Pupil position and size information can be acquired according to pupil elliptic equation.
Preferably, fitting module 400 is fitted pupil elliptic equation using least square method, and the formula of fitting is：
Wherein, A, B, C, D, E are the polynomial parameters after fitting, (xi, yi) it is the adaptive threshold fuzziness module 300
The boundary point of pupil region after segmentation, i=1 ..., m, m be pupil region boundary point number.
It in a more preferred embodiment, further include image capture module, image capture module can be video camera, such as closely
Thermal camera.
In use, facial image is acquired by image capture module, using pupil detector provided by the present embodiment,
According to the method that echelonization is handled, eyes image first is obtained by finding human eye area in facial image by adaptive lifting scheme,
The interference for eliminating Purkinje image in pupil image by morphological image process afterwards, followed by adaptive threshold fuzziness method point
Pupil is cut, finally finds out pupil elliptic equation using least square method fitting.Pupil detector operation speed provided by the invention
Degree is fast, and accuracy is high, can achieve the detection speed of 25fps under conditions of 720P resolution ratio, error is below 0.5 pixel.
Finally it should be noted that：The above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that：It still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features；
And these are modified or replaceed, technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (10)
1. a kind of pupil detection method, which is characterized in that including：
S1, according to facial image, obtain eyes image；
S2, opening operation processing is carried out to the eyes image that step S1 is obtained；
S3, the minimum gray value and average gray for calculating the eyes image after opening operation is handled, using adaptive threshold point
Segmentation method divides pupil；
S4, according to the segmentation result of step S3, be fitted pupil elliptic equation.
2. pupil detection method according to claim 1, it is characterised in that：In the step S1, adaptive boosting is utilized
Method finds human eye area by classifier, obtains eyes image.
3. pupil detection method according to claim 1, it is characterised in that：The step S1 further includes to obtained eye
Image carries out median filter process.
4. pupil detection method according to claim 1, which is characterized in that when dividing pupil in the step S3, determine
The part that eyes image gray value is greater than threshold value is set to maximum value 255 according to threshold value by threshold value, and the part less than or equal to threshold value is set
For minimum value 0.
5. pupil detection method according to claim 4, which is characterized in that the adaptive threshold fuzziness side of the step S3
In method, threshold value T is calculated using following formula：
T=Min (g (x, y))+Avg (in (x, y))/n；
Wherein, g (x, y) is by opening operation treated eyes image, and in (x, y) is the eyes image obtained in step S1,
Min (g (x, y)) is eyes image minimum gray value, and Avg (in (x, y)) is eyes image average gray, and n is the inspection of setting
Survey parameter.
6. pupil detection method according to any one of claims 1 to 5, which is characterized in that using most in the step S4
Small square law is fitted pupil elliptic equation, and the formula of fitting is：
Wherein, A, B, C, D, E are the polynomial parameters after fitting, (xi, yi) it is the pupil region that segmentation pupil obtains in step S3
Boundary point, i=1 ..., m, m be pupil region boundary point total number.
7. a kind of pupil detector, which is characterized in that including：
Ocular extraction module, for obtaining eyes image according to facial image；
Opening operation processing module, the eyes image for obtaining to ocular extraction module carry out opening operation processing；
Adaptive threshold fuzziness module, the minimum gray value and ash of treated for calculating opening operation processing module eyes image
Average value is spent, divides pupil using adaptive threshold fuzziness method；
Fitting module is fitted pupil elliptic equation for the segmentation result according to adaptive threshold fuzziness module.
8. pupil detector according to claim 7, which is characterized in that the ocular extraction module utilizes adaptive
Method for improving is answered, human eye area is found by classifier, obtains eyes image.
9. pupil detector according to claim 7, which is characterized in that the adaptive threshold fuzziness module is for true
Determine threshold value, and the part that eyes image gray value is greater than threshold value is set to by maximum value 255 according to threshold value, less than or equal to the portion of threshold value
It splits as minimum value 0；The adaptive threshold fuzziness module calculates threshold value T using following formula：
T=Min (g (x, y))+Avg (in (x, y))/n；
Wherein, g (x, y) is by the opening operation processing module treated eyes image, and in (x, y) is the ocular
The eyes image that extraction module obtains, Min (g (x, y)) are eyes image minimum gray value, and Avg (in (x, y)) is eyes image
Average gray, n are the detection parameters of setting.
10. pupil detector according to any one of claims 7 to 9, which is characterized in that the fitting module is using most
Small square law is fitted pupil elliptic equation, and the formula of fitting is：
Wherein, A, B, C, D, E are the polynomial parameters after fitting, (xi, yi) for after the adaptive threshold fuzziness module segmentation
The boundary point of pupil region, i=1 ..., m, m be pupil region boundary point total number.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810459543.8A CN108921010B (en) | 2018-05-15 | 2018-05-15 | Pupil detection method and detection device |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810459543.8A CN108921010B (en) | 2018-05-15 | 2018-05-15 | Pupil detection method and detection device |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN108921010A true CN108921010A (en) | 2018-11-30 |
| CN108921010B CN108921010B (en) | 2020-12-22 |
Family
ID=64402493
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810459543.8A Active CN108921010B (en) | 2018-05-15 | 2018-05-15 | Pupil detection method and detection device |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108921010B (en) |
Cited By (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109933200A (en) * | 2019-03-15 | 2019-06-25 | 北京环境特性研究所 | Computer vision control method based on near-infrared eyes image |
| CN110096978A (en) * | 2019-04-18 | 2019-08-06 | 温州医科大学 | The method of eccentricity cycles image procossing based on machine vision |
| CN112308014A (en) * | 2020-11-18 | 2021-02-02 | 成都集思鸣智科技有限公司 | High-speed accurate searching and positioning method for reflective points of pupils and cornea of eyes |
| CN117137427A (en) * | 2023-08-31 | 2023-12-01 | 深圳市华弘智谷科技有限公司 | Vision detection method and device based on VR and intelligent glasses |
| CN118297903A (en) * | 2024-04-08 | 2024-07-05 | 常州市新创智能科技有限公司 | Dry fiber width detecting system |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103136512A (en) * | 2013-02-04 | 2013-06-05 | 重庆市科学技术研究院 | Pupil positioning method and system |
| CN103778644A (en) * | 2014-01-15 | 2014-05-07 | 南京理工大学 | Infrared motion object detection method based on multi-scale codebook model |
| CN104517286A (en) * | 2014-12-04 | 2015-04-15 | 西安电子科技大学 | SAR (synthetic aperture radar) image registration based on self-adaption threshold segmentation and combination optimization |
| CN106203375A (en) * | 2016-07-20 | 2016-12-07 | 济南大学 | A kind of based on face in facial image with the pupil positioning method of human eye detection |
| KR101761586B1 (en) * | 2016-12-21 | 2017-07-27 | 주식회사 쓰리이 | Method for detecting borderline between iris and sclera |
| CN107844736A (en) * | 2016-09-19 | 2018-03-27 | 北京眼神科技有限公司 | iris locating method and device |
-
2018
- 2018-05-15 CN CN201810459543.8A patent/CN108921010B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103136512A (en) * | 2013-02-04 | 2013-06-05 | 重庆市科学技术研究院 | Pupil positioning method and system |
| CN103778644A (en) * | 2014-01-15 | 2014-05-07 | 南京理工大学 | Infrared motion object detection method based on multi-scale codebook model |
| CN104517286A (en) * | 2014-12-04 | 2015-04-15 | 西安电子科技大学 | SAR (synthetic aperture radar) image registration based on self-adaption threshold segmentation and combination optimization |
| CN106203375A (en) * | 2016-07-20 | 2016-12-07 | 济南大学 | A kind of based on face in facial image with the pupil positioning method of human eye detection |
| CN107844736A (en) * | 2016-09-19 | 2018-03-27 | 北京眼神科技有限公司 | iris locating method and device |
| KR101761586B1 (en) * | 2016-12-21 | 2017-07-27 | 주식회사 쓰리이 | Method for detecting borderline between iris and sclera |
Non-Patent Citations (2)
| Title |
|---|
| 彭建等: "一种基于形态学的虹膜定位快速算法", 《计算机工程与应用》 * |
| 罗玉皓等: "运用OpenCV实现红外图像的瞳孔参数实时检测", 《计算机与现代化》 * |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109933200A (en) * | 2019-03-15 | 2019-06-25 | 北京环境特性研究所 | Computer vision control method based on near-infrared eyes image |
| CN110096978A (en) * | 2019-04-18 | 2019-08-06 | 温州医科大学 | The method of eccentricity cycles image procossing based on machine vision |
| WO2020211174A1 (en) * | 2019-04-18 | 2020-10-22 | 温州医科大学 | Machine vision-based method for processing eccentric photorefraction image |
| CN112308014A (en) * | 2020-11-18 | 2021-02-02 | 成都集思鸣智科技有限公司 | High-speed accurate searching and positioning method for reflective points of pupils and cornea of eyes |
| CN112308014B (en) * | 2020-11-18 | 2024-05-14 | 成都集思鸣智科技有限公司 | High-speed accurate searching and positioning method for pupil and cornea reflecting spot of two eyes |
| CN117137427A (en) * | 2023-08-31 | 2023-12-01 | 深圳市华弘智谷科技有限公司 | Vision detection method and device based on VR and intelligent glasses |
| CN118297903A (en) * | 2024-04-08 | 2024-07-05 | 常州市新创智能科技有限公司 | Dry fiber width detecting system |
Also Published As
| Publication number | Publication date |
|---|---|
| CN108921010B (en) | 2020-12-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108921010A (en) | A kind of pupil detection method and detection device | |
| JP4307496B2 (en) | Facial part detection device and program | |
| CN106846734B (en) | A kind of fatigue driving detection device and method | |
| Gu et al. | Feature points extraction from faces | |
| JP5680667B2 (en) | System and method for identifying driver wakefulness | |
| CN108985210A (en) | A kind of Eye-controlling focus method and system based on human eye geometrical characteristic | |
| Sohail et al. | Detection of facial feature points using anthropometric face model | |
| Bhoi et al. | Template matching based eye detection in facial image | |
| JP3143819B2 (en) | Eyelid opening detector | |
| WO2010129074A1 (en) | System and method for identifying a person with reference to a sclera image | |
| JP2000259814A (en) | Image processor and method therefor | |
| KR20120006819A (en) | Gaze detection method and system adopting the same | |
| Monwar et al. | Pain recognition using artificial neural network | |
| JP4082203B2 (en) | Open / close eye determination device | |
| Rakshita | Communication through real-time video oculography using face landmark detection | |
| Kunka et al. | Non-intrusive infrared-free eye tracking method | |
| JP2004192552A (en) | Eye opening/closing determining apparatus | |
| CN109657550B (en) | Fatigue degree detection method and device | |
| KR20200049936A (en) | Biometric device and method | |
| KR101704717B1 (en) | Apparatus for recognizing iris and operating method thereof | |
| KR100397750B1 (en) | Realtime pupil detecting method for iris recognition | |
| JPH10340342A (en) | Individual identification device | |
| CN108537552B (en) | Payment method, device and system based on lens | |
| Kyaw et al. | Performance analysis of features extraction on iris recognition system | |
| CN110751064A (en) | Blink frequency analysis method and system based on image processing |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |