CN103136512A - Pupil positioning method and system - Google Patents
Pupil positioning method and system Download PDFInfo
- Publication number
- CN103136512A CN103136512A CN2013100436976A CN201310043697A CN103136512A CN 103136512 A CN103136512 A CN 103136512A CN 2013100436976 A CN2013100436976 A CN 2013100436976A CN 201310043697 A CN201310043697 A CN 201310043697A CN 103136512 A CN103136512 A CN 103136512A
- Authority
- CN
- China
- Prior art keywords
- eyes image
- pupil
- image
- pixel value
- pixels
- 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
Images
Abstract
The invention discloses a pupil positioning method and a system. The pupil positioning method comprises the following steps: acquiring eye part images, preprocessing the eye part images, increasing whole illuminance of the eye part images, and filtering and eliminating the noises of the eye part images; accounting a maximum between-class variance of all pixel points of the eye part images; based on the maximum between-class variance, computing the optimum threshold of binaryzation of the eye part images, achieving binaryzation process of the eye part images; extracting pupil edge information, and confirming the center and the radius of the pupil. The pupil positioning method can swiftly and accurately position the pupil, although under the condition of a blurring eye image, the pupil positioning method can still swiftly positioning the center of the pupil. The pupil positioning method and the system have the advantages of being simple in structure, and capable of swiftly and accurately positioning the pupil.
Description
Technical field
The present invention relates to the Visual Tracking field, be specifically related to a kind of pupil positioning method and system that can realize rapidly and accurately the pupil center location.
Background technology
The man-computer relation of exploring natural harmony has become a key areas of computer research, and natural, efficient, intelligentized human-computer interaction interface is the important trend of computer nowadays development.In field of human-computer interaction, eyes are as important information interaction passage, and sight line reaction people's attention direction, thereby line-of-sight applications is had the characteristics such as its naturality, substantivity and interactivity in field of human-computer interaction, enjoy people's concern.At present, Visual Trace Technology is ripe gradually, uses also more and more extensive.Be mainly used in the various fields such as figure/advertising research, performance analysis, scene research and man-machine interaction, in fields such as intelligent computer, intelligent appliance, virtual reality and game, good application prospect arranged also in addition.
The pupil location is as an important subject in eye tracking, and its accuracy and validity directly affect the quality of whole gaze tracking system.At present, the research method that pupil is located is a lot, and most methods is based on circle detection, mainly considers the geometrical property of pupil, often it is regarded as a circle or ellipse.Method commonly used comprises the method for Hough conversion and ellipse fitting, and with regard to its essence, Hough conversion and ellipse fitting are all to utilize the marginal information of image to obtain position and the size of pupil as the basis.The method of Hough conversion is to determine a circle by parameter search, and ellipse fitting is mainly to carry out oval match by least square method or other method.The method of Hough conversion and ellipse fitting at first all needs the eye gray level image is carried out binaryzation, then bianry image is carried out the extraction of marginal information, realizes the location with regard to its marginal information.Yet when asking for bianry image, choosing of threshold value becomes difficulty, especially is difficult for according to choosing especially of lower threshold value at infrared light.
At present, studied multiple method about the pupil location both at home and abroad, for example seek optimal value by the voting mechanism of Hough conversion at parameter space, utilize the algorithm of Hough conversion fitted ellipse to locate ellipse, utilize the Sobel operator to locate pupil in conjunction with the method that the Hough circle detects, and locate pupil based on the method that adopts Hough change detection circle in the situation at strong edge and weak edge.Although these methods have obtained certain accuracy, adopt the method for fixed threshold to increase undoubtedly the limitation of method, versatility is relatively poor, and Hough looks for round method to spread all over entire image, blind search, the method complexity is large simultaneously, resource requirement is large, and operation time is long.Just because of the deficiency of these methods, therefore all can't realize quickly and accurately the location of pupil center.
Summary of the invention
In order to overcome the defective that exists in above-mentioned prior art, the purpose of this invention is to provide a kind of pupil positioning method and system, can position pupil rapidly and accurately.
In order to realize above-mentioned purpose of the present invention, according to an aspect of the present invention, the invention provides a kind of pupil positioning method, comprise the steps:
S1: obtain eyes image;
S2: described eyes image is carried out pre-service, increase the overall brightness of described eyes image, and the noise of the described eyes image of filtering;
S3: the maximum between-cluster variance of all pixels of the described eyes image of statistics;
S4: according to described maximum between-cluster variance, calculate the optimal threshold of described eyes image binaryzation, realize the eyes image binary conversion treatment;
S5: extract pupil edge information;
S6: determine pupil center and radius.
Pupil positioning method of the present invention can position pupil rapidly and accurately, even in the situation that eye image is fuzzy, also the center of pupil can be located out fast.
In a preferred embodiment of the present invention, in described eyes image was carried out pretreated step, the method that increases the overall brightness of described eyes image was:
S21: described eyes image is converted into corresponding histogram;
S22: the gray-scale value that goes out all pixels in described eyes image according to described statistics with histogram;
S23: determine the benchmark of grey scale pixel value, the gray-scale value of the pixel of described eyes image is carried out light compensation according to described benchmark.
The present invention is by increasing the overall brightness of eyes image, also can carry out rapidly and accurately the through hole location for some vague image vegetarian refreshments or blurred picture, improved accuracy.
In another kind of preferred embodiment of the present invention, the method for the benchmark of described definite grey scale pixel value is: select the mean value of the gray-scale value of the pixel of 5%-10% in all pixels as benchmark.
The present invention selects the mean value of gray-scale value of the partial pixel point in all pixels as benchmark, and the ratio of choosing can be adjusted according to actual needs, has improved the dirigibility of method.
In a preferred embodiment of the present invention, in described eyes image is carried out pretreated step, adopt the noise of the described eyes image of median filter filtering of two dimension.
The present invention adopts the preprocess method of medium filtering and light compensation, has ability fuzzy, partially dark, promoted brightness and filtering noise by the eyes image of noise, has improved the accuracy of pupil location.
In a preferred embodiment of the present invention, in described step S3, the method for the maximum between-cluster variance of all pixels of the described eyes image of statistics is:
S31: the histogram that obtains described eyes image;
S32: determine that in described histogram in described eyes image, number of pixels is zero grey scale pixel value;
S33: determine the non-vanishing grey scale pixel value of number of pixels in described eyes image, calculate the maximum between-cluster variance of the non-vanishing grey scale pixel value of described number of pixels.
Maximum between-cluster variance computing method of the present invention only need the maximum between-cluster variance of the non-vanishing grey scale pixel value of calculating pixel number, have improved computing velocity.
In a preferred embodiment of the present invention, choose the segmentation threshold of prospect and the background of described eyes image, when described maximum between-cluster variance was maximum, described segmentation threshold was the optimal threshold of cutting apart.
The present invention utilizes follow-on maximum variance between clusters to realize that adaptive threshold extracts, and has that threshold value is suitable, a self-adaptation, fireballing advantage.
In a preferred embodiment of the present invention, the method for extracting pupil edge information is:
If the pixel value of central pixel point is 255, no matter why the pixel value of all the other 8 adjacent pixels is worth, the pixel value that keeps without exception central pixel point is 255;
If the pixel value of central pixel point is 0, and the pixel value of 8 adjacent pixels is 0, the pixel value of central pixel point changed into 255;
During all the other situations, all the pixel value with central pixel point changes 0 into.
In another kind of preferred embodiment of the present invention, utilize least square method to determine pupil center and radius.
The extracting method of pupil edge information of the present invention and based on the accurate location that the ellipse fitting method of definite pupil center of least square method and radius realizes pupil has characteristics quickly and accurately.
In order to realize above-mentioned purpose of the present invention, according to two aspects of the present invention, the invention provides a kind of pupil positioning system, comprising: eyes image acquisition module, core processing module and image display; Described eyes image acquisition module is used for gathering eyes image; Described core processing module is connected with described eyes image acquisition module, is used for receiving described eyes image and carries out according to the method described in the present invention eyes image and process and realize that pupil locates; Described image display is connected with described core processing module, is used for motion and the pupil positioning result of eyeball are shown.
Pupil positioning system of the present invention can position pupil rapidly and accurately, even in the situation that eye image is fuzzy, also the center of pupil can be located fast and is shown.
In a preferred embodiment of the present invention, also comprise data conversion module, described data conversion module is connected with described core processing module with described eyes image acquisition module respectively, and the eyes image after also changing for the format conversion that realizes described eyes image data transfers to described core processing module.
Data conversion module of the present invention carries out format conversion, when the eyes image data of the data acquisition of eyes image acquisition module collection and data type that core processing module can be processed not simultaneously, utilize this data conversion module that the data layout of eyes image acquisition module collection is changed, improved the compatibility of pupil positioning system.
In a preferred embodiment of the present invention, described eyes image acquisition module is the eye movement instrument.
The present invention adopts the eye movement instrument to gather eyes image, guarantees that eyes image is accurate, has improved accuracy and has been convenient to timely processing.
Additional aspect of the present invention and advantage part in the following description provide, and part will become obviously from the following description, or recognize by practice of the present invention.
Description of drawings
Above-mentioned and/or additional aspect of the present invention and advantage are from obviously and easily understanding becoming the description of embodiment in conjunction with following accompanying drawing, wherein:
Fig. 1 is the process flow diagram of pupil positioning method of the present invention;
Fig. 2 is the corresponding histogram of eyes image in a kind of preferred implementation of the present invention;
Fig. 3 is the eyes image figure that gathers in a kind of preferred implementation of the present invention;
Fig. 4 is the pretreating effect figure to eyes image shown in Figure 3;
Fig. 5 is the binaryzation design sketch of eyes image shown in Figure 3;
Fig. 6 is the locating effect figure of pupil center of eyes image shown in Figure 3;
Fig. 7 is the structural representation of pupil positioning system in a kind of preferred implementation of the present invention;
Fig. 8 is the structural representation of pupil positioning system in the another kind of preferred implementation 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, and wherein same or similar label represents same or similar element or the element with identical or similar functions from start to finish.Be exemplary below by the embodiment that is described with reference to the drawings, only be used for explaining the present invention, and can not be interpreted as limitation of the present invention.
In description of the invention, unless otherwise prescribed and limit, need to prove, term " installation ", " being connected ", " connection " should be done broad understanding, for example, can be mechanical connection or electrical connection, can be also the connection of two element internals, can be directly to be connected, and also can indirectly be connected by intermediary, for the ordinary skill in the art, can understand as the case may be the concrete meaning of above-mentioned term.
The invention provides a kind of pupil positioning method, as shown in Figure 1, comprise the steps:
S1: obtain eyes image;
S2: described eyes image is carried out pre-service, increase the overall brightness of described eyes image, and the noise of the described eyes image of filtering;
S3: the maximum between-cluster variance of all pixels of the described eyes image of statistics;
S4: according to described maximum between-cluster variance, calculate the optimal threshold of described eyes image binaryzation, realize the eyes image binary conversion treatment;
S5: extract pupil edge information;
S6: determine pupil center and radius.
Pupil positioning method of the present invention obtains eyes image in step S1, in a kind of preferred implementation of the present invention, gather the eye video image of eye rotation by the eye movement instrument, and be stored in computing machine; Then utilize step S2 to gather each frame eye video image to step S4, realize that the pre-service of image and self-adaption binaryzation process; Utilize at last step S5 pupil to be positioned to step S6, specifically extract pupil edge information and determine pupil center and radius.Pupil positioning method of the present invention can position pupil rapidly and accurately, even in the situation that eye image is fuzzy, also the center of pupil can be located out fast.
In a kind of preferred implementation of the present invention, the concrete steps of this pupil positioning method are:
The first step: obtain eyes image, in the present embodiment, utilize the video camera of eye movement instrument obtain the eyes image video and be stored in the storer of computing machine, Fig. 3 shows the eyes image of the three width different angles that gathered by the eye movement instrument.
Second step: eyes image is carried out pre-service, increase the overall brightness of eyes image, and the noise of filtering eyes image, image after processing as shown in Figure 4, in the present embodiment, gather each frame eyes image and do the image pre-service, increase integral image brightness and filtering noise by light compensation.In eyes image was carried out pretreated step, the method that increases the overall brightness of eyes image by light compensation was:
S21: eyes image is converted into corresponding histogram, in the present embodiment, as shown in Figure 2, in conjunction with software Visual studio2010, eyes image is converted into corresponding histogram by the image vision storehouse, its horizontal ordinate is gray-scale value, and its ordinate is the corresponding number of pixels of gray-scale value.
S22: the gray-scale value that goes out all pixels in eyes image according to statistics with histogram.
S23: determine the benchmark of grey scale pixel value, the gray-scale value benchmark of the pixel of eyes image is carried out light compensation.In the present embodiment, the method of determining the benchmark of grey scale pixel value is: select the mean value of the gray-scale value of the pixel of some in all pixels as benchmark, in an embodiment that is more preferably of the present invention, choose the mean value of gray-scale value of pixel of 5%-10% as benchmark.The mean value of the gray-scale value of the pixel of preferred employing 8% is as benchmark.The present invention selects the mean value of gray-scale value of the partial pixel point in all pixels as benchmark, and the ratio of choosing can be adjusted according to actual needs, has improved the dirigibility of method.
After determining benchmark, the gray-scale value benchmark of the pixel of eyes image is carried out light compensation.In the present embodiment, can carry out light compensation to the gray-scale value of all pixels, also can carry out light compensation to the gray-scale value of other all pixels except determining the selected pixel of benchmark.
In the present embodiment, the method of concrete light compensation can for but be not limited to the method that coefficient in proportion compensates, the certain proportion coefficient of getting the benchmark gray-scale value is added on the gray-scale value of all pixels, for example getting 50% of benchmark gray-scale value is added on the gray-scale value of all pixels, thereby realize the light compensation of the gray-scale value of all pixels has been strengthened the brightness of eyes image.The present invention is by increasing the overall brightness of eyes image, also can carry out rapidly and accurately the through hole location for some vague image vegetarian refreshments or blurred picture, improved accuracy.
After eyes image is carried out light compensation, the noise of filtering eyes image, in the present embodiment, wave filter adopts nonlinear filter, can adopt but be not limited to mean filter or and median filter, preferably adopt median filter, in a kind of embodiment that is more preferably of the present invention, adopt the noise of the median filter filtering eyes image of two dimension, concrete implementation procedure is as follows:
Slide on image with a moving window that contains odd point, the gray-scale value of window center being put the correspondence image pixel replaces with the intermediate value in window.
Because eyes image is two-dimentional, adopt two dimensional filter, this wave filter can be expressed from the next:
Wherein, S represents filter window, { f
I, jBe the sequence of image pixel, y
I, jOutput for wave filter.
The present invention adopts the preprocess method of medium filtering and light compensation, has ability fuzzy, partially dark, promoted brightness and filtering noise by the eyes image of noise, has improved the accuracy of pupil location.
The 3rd step: the maximum between-cluster variance of all pixels of statistics eyes image, in another kind of preferred implementation of the present invention, also can adopt improved maximum between-cluster variance computing method, it is not namely the maximum between-cluster variance that calculates all pixels, but calculate the maximum between-cluster variance of certain one part of pixel point, concrete method can for but be not limited to according to the eyes image histogram, count the grey scale pixel value that does not occur in eyes image, calculate the maximum between-cluster variance of remaining pixel, in the present embodiment, the method for calculating maximum between-cluster variance is:
S31: the histogram that obtains eyes image;
S32: determine that in histogram in eyes image, number of pixels is zero grey scale pixel value;
S33: determine the non-vanishing grey scale pixel value of number of pixels in eyes image, the maximum between-cluster variance of the grey scale pixel value that the calculating pixel number is non-vanishing.
Maximum between-cluster variance computing method of the present invention only need the maximum between-cluster variance of the non-vanishing grey scale pixel value of calculating pixel number, have improved computing velocity.
In a preferred embodiment of the present invention, choose the segmentation threshold of prospect and the background of eyes image, when maximum between-cluster variance was maximum, segmentation threshold was the optimal threshold of cutting apart.The present invention utilizes follow-on maximum variance between clusters to realize that adaptive threshold extracts, and has that threshold value is suitable, a self-adaptation, fireballing advantage.
In the present embodiment, after in second step, pretreated eyes image being done the histogram processing, then adopt improved maximum variance between clusters to ask for and count the grey scale pixel value that does not occur in image, only calculate the maximum between-cluster variance of remaining pixel.As shown in Figure 2, this histogram is that the corresponding number of pixels of this two part gray-scale value that reaches between 0-35 between 210-255 is all zero in the gradation of image value, therefore when asking for maximum between-cluster variance, first this two parts grey scale pixel value in image is come out, do not calculate its corresponding inter-class variance value, only the pixel of remainder is calculated.
The principle of calculating maximum between-cluster variance is as follows: meter t is the segmentation threshold of picture prospect and background, wherein to count and account for whole image scaled be w0 to prospect, its image averaging gray-scale value is u0, background is counted and accounted for image scaled is w1, its image averaging gray-scale value is u1, is u so can calculate image overall average gray scale.Traversal is t from the minimum gradation value to the maximum gradation value, and when t made inter-class variance value g maximum, this moment, t was the optimal threshold of cutting apart.
If the image size is the M*N pixel, L is the rank of gradation of image value, and in the present embodiment, the value of L is 1-255, and the gray-scale value of establishing pixel in image is denoted as N0 less than the number of pixels of t, and pixel grey scale is denoted as N1 greater than the number of pixels of t, can calculate:
N
0+N
1=M×N （3）
w
0+w
1＝1 （4）
u＝w
0u
0+w
1u
1 （5）
g＝w
0(u
0-u)
2+w
1(u
1-u)
2 （6）
Formula (5) is brought in formula (6), obtains formula (7):
g＝w
0w
1(u
0-u
1)
2 （7）
T will be from 0 to L-1 value successively, and the t value when g gets maximum is the optimal threshold of image segmentation.
The 4th step: according to maximum between-cluster variance, after the optimal threshold that the calculating eyes image is cut apart, carry out binary conversion treatment, gray-scale value is taken as p1 greater than the gray-scale value of the pixel of t value, and gray-scale value is taken as p2 less than the gray-scale value of the pixel of t value, and p1〉p2.Thereby realize the eyes image binary conversion treatment, result as shown in Figure 5.In the present embodiment, the optimal threshold of cutting apart according to the motion conditions Dynamic Acquisition eyes image of eye, and carry out dynamic binary conversion treatment according to optimal threshold.In the present embodiment, can also filter the eyes image after binary conversion treatment and make an uproar and repair, concrete method is:
extract through adaptive threshold the hole that can there be the pupil target that causes because cornea is reflective in the binaryzation eyes image that obtains, eyelash blocks the shade that brings, the noise that burr etc. causes and the situation of defective, therefore need to carry out image to the binaryzation eyes image and filter the repairing of making an uproar, in the present embodiment, adopt the corrosion operation, expansive working, opening operation and closed operation realize filtering the repairing of making an uproar, in a kind of embodiment that is more preferably of the present invention, the concrete employing corroded operation 5 times, 5 expansive workings, 1 opening operation, 1 closed operation realizes filtering preferably the repair efficiency of making an uproar, wherein, concrete function and the parameter that adopts is set to:
The corrosion handling function is: cvErode (threshold, threshold, NULL, 5);
The expansive working function is: cvDilate (threshold, threshold, NULL, 5);
The opening operation function is:
cvMorphologyEx(threshold,threshold,0,0,CV_MOP_OPEN,1);
The closed operation function is:
cvMorphologyEx(threshold,threshold,0,0,CV_MOP_CLOSE,1)。
The 5th step: extract pupil edge information on the eyes image after binary conversion treatment.In the present embodiment, the method for extraction pupil edge information is:
If the pixel value of central pixel point is 255, no matter why the pixel value of all the other 8 adjacent pixels is worth, the pixel value that keeps without exception central pixel point is 255;
If the pixel value of central pixel point is 0, and the pixel value of 8 adjacent pixels is 0, the pixel value of central pixel point changed into 255;
During all the other situations, all the pixel value with central pixel point changes 0 into.
The 6th step: determine pupil center and radius.In the present embodiment, as shown in Figure 6, utilize least square method to determine pupil center and radius.Employing realizes that based on least square method the basic thought of the match link of edge point is: ask each candidate point minimum to the quadratic sum of the upper distance of circle.If the elliptic equation of pupil is:
Ax
2+Bxy+Cy
2+Dx+Ey+F=0 （8）
Its constraint condition is:
B
2-4AC<0 （9）
Can be got by principle of least square method:
By extreme value theorem, the value of wanting to make f (A, B, C, D, E) is minimum, must have:
Can obtain thus a system of linear equations, then in conjunction with constraint condition, can solve equation coefficient A, B, C, D, E, the value of F.So oval centre coordinate (xp, yp), major axis and minor axis (a, b) all can be asked by following formula, thereby obtained radius and the centre coordinate value of pupil:
The extracting method of pupil edge information of the present invention and based on the accurate location that the ellipse fitting method of definite pupil center of least square method and radius realizes pupil has characteristics quickly and accurately.
Pupil positioning method based on the modified maximum variance between clusters of the present invention is by image pretreatment operation and the grey scale pixel value that exists in image is calculated the method for maximum between-cluster variance, fast and accurately asked for the optimal threshold of image segmentation, for the intact extraction of pupil edge information lays the foundation; Last adopt on this basis contour extraction method and realize the accurate location of pupil and improved arithmetic speed based on the ellipse fitting method of least square method.
In order to realize the location of through hole, the present invention also provides a kind of pupil positioning system, and as shown in Figure 7, it comprises eyes image acquisition module 1, core processing module 2 and image display 3.In the present embodiment, core processing module 2 is positioned at computing machine, also has storer in computing machine, be used for the eyes image that storage eyes image acquisition module 1 gathers, the user can call the eyes image in storer and check at any time by image display 3 by core processing module 2.The eyes image acquisition module 1 of this pupil positioning system is used for gathering eyes image; Core processing module 2 is connected with eyes image acquisition module 1, be used for to receive eyes image and carries out eyes image according to pupil positioning method of the present invention and process and realize that pupil locates; Image display 3 is connected with core processing module 2, is used for motion and the pupil positioning result of eyeball are shown.Pupil positioning system of the present invention can position pupil rapidly and accurately, even in the situation that eye image is fuzzy, also the center of pupil can be located fast and is shown.
In a kind of preferred implementation of the present invention, as shown in Figure 8, this pupil positioning system also comprises data conversion module 4, data conversion module 4 is connected with core processing module 2 with eyes image acquisition module 1 respectively, and the eyes image after also changing for the format conversion that realizes the eyes image data transfers to described core processing module 2.Data conversion module 4 of the present invention carries out format conversion, the eyes image data of the data acquisition that gathers when eyes image acquisition module 1 and the data type that core processing module 2 can be processed are not simultaneously, the data layout that utilizes 4 pairs of eyes image acquisition modules of this data conversion module 1 to gather is changed, and has improved the compatibility of pupil positioning system.
In the present embodiment, the eyeball view data is obtained by optical sensor, transmission via data conversion module 4 arrives in core processing module 2, completing image processes and the pupil location algorithm, and mark pupil center location, last and view data is sent image display 3 in the lump, the result that the output pupil is located.In the present embodiment, pupil positioning system position to pupil in the eye movement process shows in real time, realizes oculomotor tracking.
In the present embodiment, eyes image acquisition module 1 is eye movement instrument or optical sensor, preferably adopts the eye movement instrument, and the present invention adopts the eye movement instrument to gather eyes image, guarantees that eyes image is accurate, has improved accuracy and has been convenient to timely processing.
In the other preferred implementation of the present invention, this pupil positioning system also comprises power supply and peripheral circuit 5, be used to eyes image acquisition module 1, core processing module 2, image display 3 and data conversion module 4 that electric energy is provided, guarantee the normal operation of eyes image acquisition module 1, core processing module 2, image display 3 and data conversion module 4.
In the present embodiment, eyes image acquisition module 1 selects the MT9V011 chip of MICRON company as optical sensor SENSOR, image resolution ratio is 640*480, the frame per second of the eyes image of taking was 15 frame/seconds, was that fineness and the filming frequency of image all satisfies the usual oculomotor requirement of the mankind.Data conversion module 4 is selected CP2102, has the full speed interface of USB2.0, satisfies the large transmission requirement of data volume, and compatible with the process chip of back.Core processing module 2 is selected the ARM11 series of processes chip S3C6410 of Samsung, and speed and performance can satisfy the requirement of image and Algorithm Analysis.Image display 3 can be selected existing display module commonly used, power supply and peripheral circuit 5 can be selected existing power supply chip, as long as the operating voltage that provides satisfies the voltage request of eyes image acquisition module 1, core processing module 2, image display 3 and data conversion module 4.This pupil positioning system has good accuracy and real-time in eye movement and vision tracing process to the normal person.
In the description of this instructions, the description of reference term " embodiment ", " some embodiment ", " example ", " concrete example " or " some examples " etc. means to be contained at least one embodiment of the present invention or example in conjunction with specific features, structure, material or the characteristics of this embodiment or example description.In this manual, the schematic statement of above-mentioned term not necessarily referred to identical embodiment or example.And the specific features of description, structure, material or characteristics can be with suitable mode combinations in any one or more embodiment or example.
Although illustrated and described embodiments of the invention, those having ordinary skill in the art will appreciate that: in the situation that do not break away from principle of the present invention and aim can be carried out multiple variation, modification, replacement and modification to these embodiment, scope of the present invention is limited by claim and equivalent thereof.
Claims (10)
1. a pupil positioning method, is characterized in that, comprises the steps:
S1: obtain eyes image;
S2: described eyes image is carried out pre-service, increase the overall brightness of described eyes image, and the noise of the described eyes image of filtering;
S3: the maximum between-cluster variance of all pixels of the described eyes image of statistics;
S4: according to described maximum between-cluster variance, calculate the optimal threshold of described eyes image binaryzation, realize the eyes image binary conversion treatment;
S5: extract pupil edge information;
S6: determine pupil center and radius.
2. pupil positioning method as claimed in claim 1, is characterized in that, in described eyes image was carried out pretreated step, the method that increases the overall brightness of described eyes image was:
S21: described eyes image is converted into corresponding histogram;
S22: the gray-scale value that goes out all pixels in described eyes image according to described statistics with histogram;
S23: determine the benchmark of grey scale pixel value, the gray-scale value of the pixel of described eyes image is carried out light compensation according to described benchmark.
3. pupil positioning method as claimed in claim 2, is characterized in that, the method for the benchmark of described definite grey scale pixel value is: select the mean value of the gray-scale value of the pixel of 5%-10% in all pixels as benchmark.
4. pupil positioning method as claimed in claim 1 or 2, is characterized in that, in described eyes image is carried out pretreated step, adopts the noise of the described eyes image of median filter filtering of two dimension.
5. pupil positioning method as claimed in claim 1, is characterized in that, in described step S3, the method for the maximum between-cluster variance of all pixels of the described eyes image of statistics is:
S31: the histogram that obtains described eyes image;
S32: determine that in described histogram in described eyes image, number of pixels is zero grey scale pixel value;
S33: determine the non-vanishing grey scale pixel value of number of pixels in described eyes image, calculate the maximum between-cluster variance of the non-vanishing grey scale pixel value of described number of pixels.
6. pupil positioning method as described in claim 1 or 5, is characterized in that, chooses the segmentation threshold of prospect and the background of described eyes image, and when described maximum between-cluster variance was maximum, described segmentation threshold was the optimal threshold of cutting apart.
7. pupil positioning method as claimed in claim 1, is characterized in that, the method for extracting pupil edge information is:
If the pixel value of central pixel point is 255, no matter why the pixel value of all the other 8 adjacent pixels is worth, the pixel value that keeps without exception central pixel point is 255;
If the pixel value of central pixel point is 0, and the pixel value of 8 adjacent pixels is 0, the pixel value of central pixel point changed into 255;
During all the other situations, all the pixel value with central pixel point changes 0 into.
8. a pupil positioning system, is characterized in that, comprising: eyes image acquisition module, core processing module and image display;
Described eyes image acquisition module is used for gathering eyes image;
Described core processing module is connected with described eyes image acquisition module, is used for receiving described eyes image and carries out eyes image according to the described method of one of claim 1-7 and process and realize that pupil locates;
Described image display is connected with described core processing module, is used for motion and the pupil positioning result of eyeball are shown.
9. pupil positioning system as claimed in claim 8, it is characterized in that: also comprise data conversion module, described data conversion module is connected with described core processing module with described eyes image acquisition module respectively, and the eyes image after also changing for the format conversion that realizes described eyes image data transfers to described core processing module.
10. pupil positioning system as claimed in claim 8, it is characterized in that: described eyes image acquisition module is the eye movement instrument.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100436976A CN103136512A (en) | 2013-02-04 | 2013-02-04 | Pupil positioning method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN2013100436976A CN103136512A (en) | 2013-02-04 | 2013-02-04 | Pupil positioning method and system |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN103136512A true CN103136512A (en) | 2013-06-05 |
Family
ID=48496320
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN2013100436976A Pending CN103136512A (en) | 2013-02-04 | 2013-02-04 | Pupil positioning method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103136512A (en) |
Cited By (30)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103390152A (en) * | 2013-07-02 | 2013-11-13 | 华南理工大学 | Sight tracking system suitable for human-computer interaction and based on system on programmable chip (SOPC) |
| CN103996020A (en) * | 2014-04-10 | 2014-08-20 | 中航华东光电（上海）有限公司 | Head mounted eye tracker detection method |
| CN104463127A (en) * | 2014-12-15 | 2015-03-25 | 三峡大学 | Pupil positioning method and device |
| CN105678286A (en) * | 2016-02-29 | 2016-06-15 | 徐鹤菲 | Pupil positioning method and equipment |
| WO2016106617A1 (en) * | 2014-12-29 | 2016-07-07 | 深圳Tcl数字技术有限公司 | Eye location method and apparatus |
| CN105809128A (en) * | 2016-03-08 | 2016-07-27 | 厦门美图之家科技有限公司 | Eye pupil positioning method and system for eye part image and shooting terminal |
| CN106295489A (en) * | 2015-06-29 | 2017-01-04 | 株式会社日立制作所 | Information processing method, information processor and video monitoring system |
| CN106504271A (en) * | 2015-09-07 | 2017-03-15 | 三星电子株式会社 | Method and apparatus for eye tracking |
| CN106774863A (en) * | 2016-12-03 | 2017-05-31 | 西安中科创星科技孵化器有限公司 | A kind of method that Eye-controlling focus are realized based on pupil feature |
| CN106919933A (en) * | 2017-03-13 | 2017-07-04 | 重庆贝奥新视野医疗设备有限公司 | The method and device of Pupil diameter |
| CN108010491A (en) * | 2017-11-08 | 2018-05-08 | 孙兴珍 | A kind of background light regulation method based on pupil radium detection |
| CN108182380A (en) * | 2017-11-30 | 2018-06-19 | 天津大学 | A kind of flake pupil intelligent measurement method based on machine learning |
| WO2018108124A1 (en) * | 2016-12-15 | 2018-06-21 | 腾讯科技（深圳）有限公司 | Method and system for positioning pupil |
| CN108573204A (en) * | 2017-04-28 | 2018-09-25 | 分界线（天津）网络技术有限公司 | A kind of measurement method of pupil of human distance |
| CN108595008A (en) * | 2018-04-27 | 2018-09-28 | 北京计算机技术及应用研究所 | Man-machine interaction method based on eye movement control |
| CN108921010A (en) * | 2018-05-15 | 2018-11-30 | 北京环境特性研究所 | A kind of pupil detection method and detection device |
| CN109447945A (en) * | 2018-09-21 | 2019-03-08 | 河南农业大学 | Wheat Basic Seedling rapid counting method based on machine vision and graphics process |
| CN109766818A (en) * | 2019-01-04 | 2019-05-17 | 京东方科技集团股份有限公司 | Pupil center's localization method and system, computer equipment and readable storage medium storing program for executing |
| CN109784248A (en) * | 2019-01-02 | 2019-05-21 | 京东方科技集团股份有限公司 | Pupil positioning method, pupil positioning device, electronic equipment, storage medium |
| CN110276324A (en) * | 2019-06-27 | 2019-09-24 | 北京万里红科技股份有限公司 | The elliptical method of pupil is determined in a kind of iris image |
| CN110348408A (en) * | 2019-07-16 | 2019-10-18 | 上海博康易联感知信息技术有限公司 | Pupil positioning method and device |
| CN110751093A (en) * | 2019-10-20 | 2020-02-04 | 赵俊杰 | Method for automatically identifying pupil and calculating pupil diameter |
| CN111027502A (en) * | 2019-12-17 | 2020-04-17 | Oppo广东移动通信有限公司 | Eye image positioning method and device, electronic equipment and computer storage medium |
| CN111127537A (en) * | 2018-10-29 | 2020-05-08 | 托比股份公司 | Method and apparatus for detecting shadows in a head mounted device |
| WO2020211174A1 (en) * | 2019-04-18 | 2020-10-22 | 温州医科大学 | Machine vision-based method for processing eccentric photorefraction image |
| CN112258569A (en) * | 2020-09-21 | 2021-01-22 | 苏州唐古光电科技有限公司 | Pupil center positioning method, device, equipment and computer storage medium |
| CN112989878A (en) * | 2019-12-13 | 2021-06-18 | Oppo广东移动通信有限公司 | Pupil detection method and related product |
| CN113509138A (en) * | 2021-05-11 | 2021-10-19 | 岭南师范学院 | Laser self-mixing interference type autism child pupil light reflectometer |
| CN114020155A (en) * | 2021-11-05 | 2022-02-08 | 沈阳飞机设计研究所扬州协同创新研究院有限公司 | High-precision sight line positioning method based on eye tracker |
| WO2022257120A1 (en) * | 2021-06-11 | 2022-12-15 | 华为技术有限公司 | Pupil position determination method, device and system |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100284576A1 (en) * | 2006-09-25 | 2010-11-11 | Yasunari Tosa | Iris data extraction |
-
2013
- 2013-02-04 CN CN2013100436976A patent/CN103136512A/en active Pending
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20100284576A1 (en) * | 2006-09-25 | 2010-11-11 | Yasunari Tosa | Iris data extraction |
Non-Patent Citations (2)
| Title |
|---|
| 黄丽丽 等: "基于改进型最大类间方差法的瞳孔定位方法", 《万方学术期刊数据库》 * |
| 黄祎: "虹膜识别的预处理算法研究", 《CNKI中国优秀硕士学位论文全文数据库》 * |
Cited By (41)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103390152B (en) * | 2013-07-02 | 2017-02-08 | 华南理工大学 | Sight tracking system suitable for human-computer interaction and based on system on programmable chip (SOPC) |
| CN103390152A (en) * | 2013-07-02 | 2013-11-13 | 华南理工大学 | Sight tracking system suitable for human-computer interaction and based on system on programmable chip (SOPC) |
| CN103996020A (en) * | 2014-04-10 | 2014-08-20 | 中航华东光电（上海）有限公司 | Head mounted eye tracker detection method |
| CN104463127A (en) * | 2014-12-15 | 2015-03-25 | 三峡大学 | Pupil positioning method and device |
| WO2016106617A1 (en) * | 2014-12-29 | 2016-07-07 | 深圳Tcl数字技术有限公司 | Eye location method and apparatus |
| CN106295489B (en) * | 2015-06-29 | 2021-09-28 | 株式会社日立制作所 | Information processing method, information processing device and video monitoring system |
| CN106295489A (en) * | 2015-06-29 | 2017-01-04 | 株式会社日立制作所 | Information processing method, information processor and video monitoring system |
| CN106504271A (en) * | 2015-09-07 | 2017-03-15 | 三星电子株式会社 | Method and apparatus for eye tracking |
| CN106504271B (en) * | 2015-09-07 | 2022-01-25 | 三星电子株式会社 | Method and apparatus for eye tracking |
| CN105678286B (en) * | 2016-02-29 | 2019-03-15 | 徐鹤菲 | A kind of pupil positioning method and equipment |
| CN105678286A (en) * | 2016-02-29 | 2016-06-15 | 徐鹤菲 | Pupil positioning method and equipment |
| CN105809128A (en) * | 2016-03-08 | 2016-07-27 | 厦门美图之家科技有限公司 | Eye pupil positioning method and system for eye part image and shooting terminal |
| CN106774863A (en) * | 2016-12-03 | 2017-05-31 | 西安中科创星科技孵化器有限公司 | A kind of method that Eye-controlling focus are realized based on pupil feature |
| US10878593B2 (en) | 2016-12-15 | 2020-12-29 | Tencent Technology (Shenzhen) Company Limited | Pupil localizing method and system |
| WO2018108124A1 (en) * | 2016-12-15 | 2018-06-21 | 腾讯科技（深圳）有限公司 | Method and system for positioning pupil |
| CN106919933A (en) * | 2017-03-13 | 2017-07-04 | 重庆贝奥新视野医疗设备有限公司 | The method and device of Pupil diameter |
| CN108573204A (en) * | 2017-04-28 | 2018-09-25 | 分界线（天津）网络技术有限公司 | A kind of measurement method of pupil of human distance |
| CN108010491A (en) * | 2017-11-08 | 2018-05-08 | 孙兴珍 | A kind of background light regulation method based on pupil radium detection |
| CN108010491B (en) * | 2017-11-08 | 2018-09-07 | 深圳市星王电子有限公司 | A kind of background light regulation method based on pupil radium detection |
| CN108182380A (en) * | 2017-11-30 | 2018-06-19 | 天津大学 | A kind of flake pupil intelligent measurement method based on machine learning |
| CN108182380B (en) * | 2017-11-30 | 2023-06-06 | 天津大学 | Intelligent fisheye pupil measurement method based on machine learning |
| CN108595008B (en) * | 2018-04-27 | 2022-02-08 | 北京计算机技术及应用研究所 | Human-computer interaction method based on eye movement control |
| CN108595008A (en) * | 2018-04-27 | 2018-09-28 | 北京计算机技术及应用研究所 | Man-machine interaction method based on eye movement control |
| CN108921010A (en) * | 2018-05-15 | 2018-11-30 | 北京环境特性研究所 | A kind of pupil detection method and detection device |
| CN109447945A (en) * | 2018-09-21 | 2019-03-08 | 河南农业大学 | Wheat Basic Seedling rapid counting method based on machine vision and graphics process |
| CN111127537A (en) * | 2018-10-29 | 2020-05-08 | 托比股份公司 | Method and apparatus for detecting shadows in a head mounted device |
| CN109784248A (en) * | 2019-01-02 | 2019-05-21 | 京东方科技集团股份有限公司 | Pupil positioning method, pupil positioning device, electronic equipment, storage medium |
| CN109766818A (en) * | 2019-01-04 | 2019-05-17 | 京东方科技集团股份有限公司 | Pupil center's localization method and system, computer equipment and readable storage medium storing program for executing |
| WO2020211174A1 (en) * | 2019-04-18 | 2020-10-22 | 温州医科大学 | Machine vision-based method for processing eccentric photorefraction image |
| CN110276324A (en) * | 2019-06-27 | 2019-09-24 | 北京万里红科技股份有限公司 | The elliptical method of pupil is determined in a kind of iris image |
| CN110276324B (en) * | 2019-06-27 | 2021-06-22 | 北京万里红科技股份有限公司 | Method for determining pupil ellipse in iris image |
| CN110348408A (en) * | 2019-07-16 | 2019-10-18 | 上海博康易联感知信息技术有限公司 | Pupil positioning method and device |
| CN110751093A (en) * | 2019-10-20 | 2020-02-04 | 赵俊杰 | Method for automatically identifying pupil and calculating pupil diameter |
| CN110751093B (en) * | 2019-10-20 | 2023-10-27 | 赵俊杰 | Method for automatically identifying pupil and calculating pupil diameter |
| CN112989878A (en) * | 2019-12-13 | 2021-06-18 | Oppo广东移动通信有限公司 | Pupil detection method and related product |
| CN111027502A (en) * | 2019-12-17 | 2020-04-17 | Oppo广东移动通信有限公司 | Eye image positioning method and device, electronic equipment and computer storage medium |
| CN112258569A (en) * | 2020-09-21 | 2021-01-22 | 苏州唐古光电科技有限公司 | Pupil center positioning method, device, equipment and computer storage medium |
| CN112258569B (en) * | 2020-09-21 | 2024-04-09 | 无锡唐古半导体有限公司 | Pupil center positioning method, pupil center positioning device, pupil center positioning equipment and computer storage medium |
| CN113509138A (en) * | 2021-05-11 | 2021-10-19 | 岭南师范学院 | Laser self-mixing interference type autism child pupil light reflectometer |
| WO2022257120A1 (en) * | 2021-06-11 | 2022-12-15 | 华为技术有限公司 | Pupil position determination method, device and system |
| CN114020155A (en) * | 2021-11-05 | 2022-02-08 | 沈阳飞机设计研究所扬州协同创新研究院有限公司 | High-precision sight line positioning method based on eye tracker |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103136512A (en) | Pupil positioning method and system | |
| CN103176607A (en) | Eye-controlled mouse realization method and system | |
| CN102360421B (en) | Face identification method and system based on video streaming | |
| CN106778664B (en) | Iris image iris area segmentation method and device | |
| CN104143079B (en) | The method and system of face character identification | |
| CN107463931A (en) | A kind of real-time pointer instrument reading method and device based on ARM platforms | |
| CN104361314A (en) | Method and device for positioning power transformation equipment on basis of infrared and visible image fusion | |
| CN102878647B (en) | Automatic monitoring system and automatic monitoring method for operating state of air conditioner | |
| CN103295016A (en) | Behavior recognition method based on depth and RGB information and multi-scale and multidirectional rank and level characteristics | |
| CN104615978A (en) | Sight direction tracking method and device | |
| CN103438834B (en) | The hierarchical quick three-dimensional measurement mechanism of structure based light projection and measuring method | |
| CN106214118A (en) | A kind of ocular movement based on virtual reality monitoring system | |
| CN104571474A (en) | Method and device for adaptively adjusting contents displayed on terminal screen | |
| CN110287791A (en) | A kind of screening technique and system for face picture | |
| CN106303296A (en) | A kind of image mosaic emerging system | |
| CN105872351A (en) | Method and device for shooting picture in backlight scene | |
| CN103473776A (en) | Method and system for comparing image definition and automatic focusing control method | |
| CN102025919A (en) | Method and device for detecting image flicker and camera applying device | |
| CN104966300A (en) | Bearing roller image detection system, method and image detection device | |
| CN103440492A (en) | Hand-held cigarette recognizer | |
| CN106915303A (en) | Automobile A-column blind area perspective method based on depth data and fish eye images | |
| CN114881956A (en) | Cell panel grid line self-adaptive extraction method and system based on Hough transform | |
| CN103605171A (en) | All-sky imaging instrument and cloud layer characteristic analysis method based on all-sky imaging instrument | |
| CN102999890A (en) | Method for dynamically calibrating image light intensity on basis of environmental factors | |
| CN101196984B (en) | Fast face detecting method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20130605 |