CN103632137A - Human iris image segmentation method - Google Patents
Human iris image segmentation method Download PDFInfo
- Publication number
- CN103632137A CN103632137A CN201310570892.4A CN201310570892A CN103632137A CN 103632137 A CN103632137 A CN 103632137A CN 201310570892 A CN201310570892 A CN 201310570892A CN 103632137 A CN103632137 A CN 103632137A
- Authority
- CN
- China
- Prior art keywords
- human eye
- iris
- image
- training set
- canthus
- 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
Images
Landscapes
- Measurement Of The Respiration, Hearing Ability, Form, And Blood Characteristics Of Living Organisms (AREA)
- Image Analysis (AREA)
Abstract
The invention relates t o a human iris image segmentation method. The method comprises the following steps of step 1: gradually locating canthus of human eye for a training set of human eye image samples one by one; step 2: gradually locating the center of human eyes for a training set of human iris image samples one by one; step 3: batch aligning the training sets by adopting an algorithm on the basis of sparsity and low-rank decomposition; step 4: segmenting the iris of the training set of human eye images for the batch-aligned training set by adopting a method integrating the canny margin detection and hough transformation; step 5: carrying out the iris segmentation for an inputted test picture. The algorithm based on the sparsity and low-rank decomposition is used for batch aligning the training set of samples, so that the problem in a great number of samples such as inconsistence in brightness and shading of eyelashes can be solved, a purpose for realizing the iris segmentation on the image with brightness fluctuation and shading problem being eliminated and the test image is realized by adopting the canny margin detection and hough transformation. The method is widely applied to the iris recognition field.
Description
Technical field
The present invention relates to image model identification field, particularly a kind of human eye iris segmentation method.
Background technology
We are in the society of an advanced IT application at present, people are increasing to information requirement, meanwhile, also more and more higher to the security requirement of information, and identity recognizing technology is exactly a method that improves Information Security, it is subject to people's great attention in increasing field.Identification is some uniqueness characteristic that utilizes human body, adopts some technology to differentiate this this feature, thereby people's identity is identified.Previous conventional identity recognizing technology relies on the features such as people's face, fingerprint, hand-type, signature to identify, but these features are all the surfaces of some human bodies, exist very large easy change, make to rely on these features to carry out identification meeting and become and be not very reliable.In recent years, risen iris recognition technology, due to the uniqueness of iris, unchangeable property, can not, by uniquenesses such as operation change, make it to have played more and more important effect in scientific research and industrial circle.But, special construction due to iris, in image acquisition process, we can not shoot pure iris image, conventionally in the iris image collecting, not only comprise iris, also comprise other parts of eyes, as pupil, eyelid, eyelashes etc., iris recognition technology can not directly be identified such picture, can only identify iris portion.Therefore an important pre-service of iris recognition technology is exactly human eye iris segmentation.
The result of human eye iris segmentation is normally for iris identification, it is the direct objective for implementation of iris identity recognizing technology, the accuracy of cutting apart badly influences the accuracy of identification, therefore iris splitting method is very important, is to ensure iris recognition one of preprocessing means and gordian technique accurately.
Human eye iris segmentation is mainly directly eye image to be detected, and is partitioned into iris portion wherein.Iris segmentation method mainly utilizes iris outer edge to be approximately circular, adopts the circle detection method in iris outer edge modeling or image to carry out.Wherein, following steps are often followed in the modeling of iris outer edge: rim detection and edge modeling.Rim detection is normally carried out rim detection to the imagery exploitation canny after gaussian filtering or sobel operator.Then the edge that normally the method edge by mathematics detects in the image of binaryzation carries out modeling.And circle detection in image is also first to adopt rim detection conventionally, then the image of edge after detecting adopt Hough to change the inside and outside circle border of detecting iris, thereby realize cutting apart of iris.
Although above-mentioned these algorithm comparative maturities, these methods are all existing identical shortcoming.Owing to existing upper and lower ciliary the blocking of brightness variation and human eye in the iris sample normally collecting, can make the circle of inner and outer boundary of iris not obvious, cause segmentation errors.In the iris image being simultaneously partitioned into, exist the shortcomings such as brightness variation, human eye upper and lower are ciliaryly blocked, sample image does not align, can cause on follow-up iris recognition very large impact, the accuracy of identification is reduced.
Summary of the invention
The invention provides a kind of human eye iris segmentation method, eye image to be detected in batches can align, remove ciliary blocking in the brightness variation in eye image, upper lower eyelid, set up clear, go to block, the eye image sample pattern of alignment in batches, then these samples are realized to iris segmentation.
The technical scheme principal character that the present invention addresses the above problem is: its concrete steps are as follows:
Step 1: the training set eye image sample to key words sorting carries out location, human eye canthus one by one, utilize Harris Corner Detection Algorithm to detect the angle point in eye image, then these angle points are retrieved to traversal, selecting the point of horizontal ordinate minimum is human eye canthus;
Step 2: training set human eye iris image sample is located to human eye center one by one, first by threshold value, samples pictures to be detected is carried out to binaryzation, then utilize canny edge detection method to carry out to binary image the boundary graph that rim detection obtains binaryzation, recycling Hough transformation detects round to boundary graph, the circle detecting is just decided to be the inner edge of human eye iris, and human eye center is elected at Jiang Ciyuan center as;
Step 3: utilize the human eye canthus point and the human eye center that detect in training set, adopt (the Robust Alignment by Sparse and Low-rank Decomposition) algorithm based on sparse and low-rank decomposition to carry out batch alignment to training set sample;
Step 4: the training set after aliging is in batches cut apart, first by canny edge detecting technology, human eye iris image to be detected is converted into the boundary graph of binaryzation, utilize Hough transformation in boundary graph, to find the circle of radius between pupil maximum radius and iris maximum radius, the circle drawing is defined as to iris external diameter circle, and then continue to utilize Hough transformation to find another circle in the iris external diameter circle region of finding out, the circle obtaining is defined as to iris internal diameter circle, so just the region between iris external diameter circle and iris internal diameter circle can be decided to be to iris region, thereby realize cutting apart training set iris,
Step 5: eye image repeating step one and step 2 to input, find canthus and the human eye center of the eye image of input, recycling canthus and the oculocentric line of people rotate eye image, make this line be adjusted to horizontal level, finally, to the image repeating step four after adjusting, realize the human eye iris segmentation to input picture.
The invention has the beneficial effects as follows: of the present invention is a kind of human eye iris segmentation method, the method has first been chosen automatically based on sparse and two reference points low-rank decomposition algorithm, utilization realizes training sample based on sparse and low-rank decomposition algorithm and aligns in batches, there is good denoising effect, the brightness that can remove in training set changes, ciliary blocking in upper lower eyelid; By test sample book is rotated, test sample book is alignd, the quality of the sample image that this not only improves, and the accuracy that can improve follow-up iris recognition with training sample.Therefore, the method can be widely used in iris recognition field.
On the basis of technique scheme, the present invention can also do following improvement.
Further, described human eye canthus detects and utilizes Harris Corner Detection to carry out.Concrete steps are as follows:
1). adopt the directional derivative of canny operator computed image, calculate respectively the directional derivative in horizontal direction and vertical direction;
2). calculate the coefficient correlation matrix of each pixel;
3). calculate the angle point value of each pixel;
4). find out maximal value in all angle point values;
5). travel through all angle point values, if detect the angle point value of pixel, be greater than 0.01 times of maximum angular point value, and be maximal value in specifying neighborhood pixels region, just this pixel is labeled as to angle point;
6). all angle points are traveled through, and the leftmost position according to human eye canthus in human eye, is labeled as human eye canthus by the angle point of horizontal ordinate minimum.
Further, described human eye Spot detection is based on canny rim detection and Hough transformation.Concrete steps are as follows:
1). picture to be split is carried out to binaryzation based on threshold value, obtain binary image;
2). binary image is carried out to smothing filtering with Gaussian filter, use amplitude and the direction of canny operator calculation of filtered image gradient;
3). along the gradient direction calculating, detect, the pixel that is not local maximum is set to 0, gradient direction is carried out to non-maximum value inhibition, obtain the edge that image only has a pixel width;
4). choose two threshold values
th1with
th2(
), non-maximum value inhibition image is processed and obtained two width images.Image 1 is less than Grad
th1the gray-scale value of pixel be made as 0, the pixel value that is greater than threshold value is constant; Image 2 is less than Grad
th2the gray-scale value of pixel be made as 0, the pixel value that is greater than threshold value is constant.Image 2 is scanned, when running into the pixel of a non-zero gray scale
p (x, y)time, follow the tracks of with
p (x, y)for the outline line of starting point, until the terminal of outline line
q (x, y).In image under consideration 1 with image 2 in
q (x, y)point corresponding to some position
s (x, y)8 adjacent domains.If
s (x, y)in 8 adjacent domains of point, there is non-zero pixels
s1 (x, y)exist, be included in image 2, as
r (x, y)point.With
r (x, y)for starting point, repeat the scanning to image 2, until to all continuing in image 1 and image 2.When completing comprising
p (x, y)the link of outline line after, this outline line is labeled as and is accessed.Continuation, to image 2 scannings, is found next outline line.Repeat track, until can not find new outline line in image 2;
5). set up parameter space
a (a, b, r), from first pixel of image 2, make
aequal the horizontal ordinate of current pixel,
bequal the ordinate of current pixel,
rinitial value be made as predefined minimum value, find in image 2 through with
(a, b)for the center of circle,
rnon-zero pixel number for radius, is kept at
a (a, b, r)in.Then make
radd 1, repeat to find step, and preserve corresponding result, until
requal predefined maximal value.Complete to current with
(a, b)for after the circle in the center of circle searches, make
(a, b)for the next point of current pixel, repeating step, until
(a, b)last pixel for image, completes parameter space
a (a, b, r)assignment;
6). determine parameter space
amaximal value, maximal value is corresponding
(a, b)be the circular center of circle detecting, this point is labeled as to human eye center;
7). re-execute step 1 and two, until all training set samples have all carried out the detection at human eye canthus and center.
Further, the alignment of the batch of described training set sample is based on sparse and low-rank decomposition algorithm.Concrete steps are as follows:
1). according to
to matrix
dcarrying out low-rank decomposition obtains
awith
e.Wherein,
drepresent to belong to of a sort training set sample matrix to be split, each column vector of matrix represents a sample, and its element is that each pixel value in this samples pictures is arranged in order and forms;
aimage pattern result after expression is alignd in batches to training set,
abe one with
dthe matrix that size is identical, each column vector is illustrated in
dthe sample that the column vector at middle same position place represents goes the result forming after dry and alignment;
represent a weight parameter, be appointed as
(
nthe number that represents sample in this class training set).
expression is to sample matrix
da conversion, wherein 2, human eye canthus and center are for initialization
,
ebe one with
awith
dthe matrix that size is identical,
eeach column vector be illustrated in
dbrightness in the samples pictures of the column vector at middle same position place changes and blocks;
representing matrix
e0 normal form;
2). according to
a, each column vector is reduced into a sheet by a sheet image, each pixel value of image equals each element of current column vector successively, obtains the image pattern after alignment in batches.
Further, the process that the described training set to after aliging is in batches realized human eye iris segmentation is first to adopt canny rim detection, adopt again Hough transformation first to detect the circle of a radius between maximum pupil and maximum iris radius, be defined as iris external diameter, more further to the region in iris external diameter circle, adopt Hough variation to carry out iris inner diameter measurement.
accompanying drawing explanation
Fig. 1 is a kind of human eye iris segmentation method general flow chart the present invention relates to;
Fig. 2 is the human eye canthus detection method flow chart of steps the present invention relates to;
Fig. 3 is the human eye central point detection method flow chart of steps the present invention relates to;
Fig. 4 is that use canny rim detection and the Hough transformation that the present invention relates to are realized iris segmentation flow chart of steps.
Embodiment
Below in conjunction with accompanying drawing, principle of the present invention and feature are described, example, only for explaining the present invention, is not intended to limit scope of the present invention.
Fig. 1 is a kind of human eye iris segmentation method general flow chart the present invention relates to; Fig. 2 is the human eye canthus detection method flow chart of steps the present invention relates to; Fig. 3 is the human eye central point detection method flow chart of steps the present invention relates to; Fig. 4 is that use canny rim detection and the Hough transformation that the present invention relates to are realized iris segmentation flow chart of steps; Press shown in Fig. 1,2,3,4, a kind of typical implementation step of the present invention is as follows:
By user, provide image to be split.
Step 1: the training set human eye iris image sample to known classification carries out location, human eye canthus: utilize Harris Corner Detection Algorithm to detect one by one the angle point in eye image, and these angle points are carried out to full search spread, select the angle point of horizontal ordinate minimum to orientate position, human eye canthus as.Concrete steps are as follows:
1). adopt the directional derivative of canny operator computed image, calculate respectively the directional derivative in horizontal direction and vertical direction
with
;
2). calculate the coefficient correlation matrix of each pixel
, wherein
representative and Gauss's template
do convolution, the variance of this template is 2, and neighborhood window size is 7*7;
3). calculate the angle point value of each pixel
;
represent to ask matrix
determinant,
representing matrix
mark,
represent weighted value, between 0.04 to 0.06;
4). find out all angle point values
middle maximal value
;
5). travel through all angle point values, if detect the angle point value of pixel, be greater than 0.01
, and be maximal value in specifying neighborhood pixels region, just this pixel being labeled as to angle point, the size in neighborhood pixels region is 9*9;
6). all angle points are traveled through, and the leftmost position according to human eye canthus in human eye, is labeled as human eye canthus by the angle point of horizontal ordinate minimum.
Step 2: the human eye center based on canny rim detection and Hough transformation detection training set sample.Concrete steps are as follows:
1). image to be split is carried out to binaryzation based on threshold value, obtain binary image, the size of threshold value is appointed as 20;
2). binary image is carried out to smothing filtering with Gaussian filter, use amplitude and the direction of canny operator calculation of filtered image gradient;
3). along the gradient direction calculating, detect, the pixel that is not local maximum is set to 0, gradient direction is carried out to non-maximum value inhibition, obtain the edge that image only has a pixel width;
4). choose two threshold values
th1with
th2(
), non-maximum value inhibition image is processed and obtained two width images.Image 1 is less than Grad
th1the gray-scale value of pixel be made as 0, the pixel value that is greater than threshold value is constant; Image 2 is less than Grad
th2the gray-scale value of pixel be made as 0, the pixel value that is greater than threshold value is constant.Image 2 is scanned, when running into the pixel of a non-zero gray scale
p (x, y)time, follow the tracks of with
p (x, y)for the outline line of starting point, until the terminal of outline line
q (x, y).In image under consideration 1 with image 2 in
q (x, y)point corresponding to some position
s (x, y)8 adjacent domains.If
s (x, y)in 8 adjacent domains of point, there is non-zero pixels
s1 (x, y)exist, be included in image 2, as
r (x, y)point.With
r (x, y)for starting point, repeat the scanning to image 2, until to all continuing in image 1 and image 2.When completing comprising
p (x, y)the link of outline line after, this outline line is labeled as and is accessed.Continuation, to image 2 scannings, is found next outline line.Repeat track, until can not find new outline line in image 2;
5). set up parameter space
a (a, b, r), from first pixel of image 2, make a equal the horizontal ordinate of current pixel, b equals the ordinate of current pixel, the initial value of r is made as predefined minimum value, finds in image 2 through take (a, b) as the center of circle, the non-zero pixel number that r is radius, be kept in A (a, b, r).Then make r add 1, repeat to find step, and preserve corresponding result, until r equals predefined maximal value.Complete and take after circle that (a, b) be the center of circle searches to current, make (a, b) for the next point of current pixel, repeating step, until last pixel that (a, b) is image completes the assignment to parameter space A (a, b, r);
6). determine the maximal value of parameter space A, corresponding (a, the b) of maximal value is the circular center of circle detecting, and this point is labeled as to human eye center;
7). re-execute step 1 and two, until all training set samples have all carried out the detection at human eye canthus and center.
Step 3: utilize human eye canthus and center as initial transformation reference point at 2, adopt and carry out batch based on sparse and algorithm (Robust Alignment by Sparse and Low-rank Decomposition, RASL) low-rank decomposition and align.Concrete steps are as follows:
1). according to
to matrix
dcarrying out low-rank decomposition obtains
awith
e.
drepresent to belong to of a sort training set sample matrix to be split, each column vector of matrix represents a sample, and its element is that each pixel value in this samples pictures is arranged in order and forms.
aimage pattern result after expression is alignd in batches to training set,
abe one with
dthe matrix that size is identical, each column vector is illustrated in
dthe sample that the column vector at middle same position place represents goes the result forming after dry and alignment.
represent a weight parameter, be appointed as
(
nthe number that represents sample in this class training set).
expression is to sample matrix
da conversion, wherein 2, human eye canthus and center are for initialization
,
ebe one with
awith
dthe matrix that size is identical,
eeach column vector be illustrated in
dbrightness in the samples pictures of the column vector at middle same position place changes and blocks,
representing matrix
e0 normal form;
2). according to
a, each column vector is reduced into a sheet by a sheet image, each pixel value of image equals each element of current column vector successively, obtains the image pattern after alignment in batches.
Step 4: the image after aliging is in batches realized to human eye iris segmentation.Concrete steps are as follows:
1). to the sample after alignment in batches adopt step 2 2), 3), 4), 5) obtain the outline map of binaryzation;
2). determine the maximal value of parameter space A, draw the circle that parameter (a, b, r) is corresponding, this circle is human eye iris external diameter;
3). determine iris external diameter, in the region of iris external diameter 5 of repeating step two);
4). determine the maximal value of parameter space A, draw the circle that parameter (a, b, r) is corresponding, this circle is human eye iris internal diameter, and the region between iris external diameter and internal diameter is decided to be to iris region, so just can realize cutting apart of human eye iris;
5). repeat 1), 2), 3), 4), 5), until all training samples all segment.
Step 5: eye image to be identified (test sample book) repeating step one and step 2 to input, find canthus and the human eye center of the eye image of input, recycling canthus and the oculocentric line of people rotate eye image, make this line be adjusted to horizontal level, thereby the eye image of input is alignd with training set, finally, to the image repeating step four after adjusting, realize the human eye iris segmentation to input picture.
The foregoing is only preferred embodiment of the present invention, in order to limit the present invention, within the spirit and principles in the present invention not all, any modification of doing, be equal to replacement, improvement etc., within all should being included in protection scope of the present invention.
Claims (7)
1. a human eye iris segmentation method, is characterized in that, the method concrete steps are as follows:
Step 1: training set eye image sample is carried out to location, human eye canthus one by one, utilize Harris Corner Detection Algorithm to detect the angle point in eye image, then these angle points are retrieved to traversal, selecting the point of horizontal ordinate minimum is human eye canthus;
Step 2: training set human eye iris image sample is located to human eye center one by one, first by threshold value, samples pictures to be detected is carried out to binaryzation, then utilize canny edge detection method to carry out to binary image the boundary graph that rim detection obtains binaryzation, recycling Hough transformation detects round to boundary graph, the circle detecting is just decided to be the inner edge of human eye iris, and human eye center is elected at Jiang Ciyuan center as;
Step 3: utilize the human eye canthus point and the human eye center that detect in training set, adopt (the Robust Alignment by Sparse and Low-rank Decomposition) algorithm based on sparse and low-rank decomposition to carry out batch alignment to training set human eye iris image sample;
Step 4: the training set human eye iris image after aliging is in batches cut apart, first by Canny edge detecting technology, human eye iris image to be detected is converted into the boundary graph of binaryzation, utilize Hough transformation in boundary graph, to find the circle of radius between pupil maximum radius and iris maximum radius, the circle drawing is defined as to iris external diameter circle, and then continue to utilize Hough transformation to find another circle in the iris external diameter circle region of finding out, the circle obtaining is defined as to iris internal diameter circle, so just the region between iris external diameter circle and iris internal diameter circle can be decided to be to iris region, thereby realize cutting apart iris in training set image,
Step 5: eye image repeating step one and step 2 to input, find canthus and the human eye center of the eye image of input, recycling canthus and the oculocentric line of people rotate eye image, make this line be adjusted to horizontal level, finally, to the image repeating step four after adjusting, realize the human eye iris segmentation to input picture.
2. a kind of human eye iris segmentation method according to claim 1, is characterized in that, the location at human eye canthus utilizes human eye canthus to carry out in the leftmost positional information of human eye.
3. a kind of human eye iris segmentation method according to claim 1, it is characterized in that, the location at human eye canthus is to be undertaken by angle point, detected angle point is carried out to full search spread, when the horizontal ordinate of angle point to be compared is less than the horizontal ordinate of the angle point having detected, current angle point is labeled as to position, human eye canthus, travels through until all angle points have traveled through entirely successively.
4. a kind of human eye iris segmentation method according to claim 1, is characterized in that, described two reference points based in sparse and low-rank decomposition algorithm are calculated by step 1 claimed in claim 1 and step 2, rather than manually choose.
5. a kind of human eye iris segmentation method according to claim 1, it is characterized in that, utilizing canny rim detection and Hough transformation to carry out before human eye iris internal-and external diameter circle detects, the batch alignment of training sample is carried out in utilization based on sparse and algorithm low-rank decomposition, removed the interference of the factor of blocking of eyelash and eyelid in brightness variation, upper lower eyelid, reduce the brightness in sample to be split and blocked the interference of unfavorable factor, improved the quality of training set sample.
6. a kind of human eye iris segmentation method according to claim 1, it is characterized in that, training set sample is adopted based on sparse and algorithm low-rank decomposition and can ajust the position of human eye in image pattern, thereby reduced in image acquisition process collecting device with respect to the inconsistent interference in acquisition target position.
7. a kind of human eye iris segmentation method according to claim 1, it is characterized in that, test sample book is rotated, can ajust the position of human eye in image pattern in test sample book, thereby test sample book is alignd with training set sample, improved the quality of test sample book.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310570892.4A CN103632137B (en) | 2013-11-15 | 2013-11-15 | A kind of human eye iris segmentation method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201310570892.4A CN103632137B (en) | 2013-11-15 | 2013-11-15 | A kind of human eye iris segmentation method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN103632137A true CN103632137A (en) | 2014-03-12 |
| CN103632137B CN103632137B (en) | 2016-08-24 |
Family
ID=50213166
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201310570892.4A Expired - Fee Related CN103632137B (en) | 2013-11-15 | 2013-11-15 | A kind of human eye iris segmentation method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN103632137B (en) |
Cited By (9)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104573660A (en) * | 2015-01-13 | 2015-04-29 | 青岛大学 | Method for precisely positioning human eyes by SIFT point descriptor |
| CN109165306A (en) * | 2018-08-09 | 2019-01-08 | 长沙理工大学 | Image search method based on the study of multitask Hash |
| CN109376649A (en) * | 2018-10-20 | 2019-02-22 | 张彦龙 | A method of likelihood figure, which is reduced, from eye gray level image calculates the upper lower eyelid of identification |
| WO2019104616A1 (en) * | 2017-11-30 | 2019-06-06 | 深圳配天智能技术研究院有限公司 | Image processing method and device and computer-readable storage medium |
| CN110022753A (en) * | 2016-09-29 | 2019-07-16 | 奇跃公司 | Neural network for eye image segmentation and picture quality estimation |
| CN110619273A (en) * | 2019-08-14 | 2019-12-27 | 中山市奥珀金属制品有限公司 | Efficient iris recognition method and device |
| CN110781745A (en) * | 2019-09-23 | 2020-02-11 | 杭州电子科技大学 | Tail eyelash detection method based on composite window and gradient weighted direction filtering |
| CN111144413A (en) * | 2019-12-30 | 2020-05-12 | 福建天晴数码有限公司 | Iris positioning method and computer readable storage medium |
| CN111241951A (en) * | 2020-01-03 | 2020-06-05 | 中山市奥珀金属制品有限公司 | Iris image processing method and device |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060280361A1 (en) * | 2005-06-14 | 2006-12-14 | Canon Kabushiki Kaisha | Image processing apparatus, image processing method,computer program, and storage medium |
| CN1900951A (en) * | 2006-06-02 | 2007-01-24 | 哈尔滨工业大学 | Iris image flexible specification method based on mathematical morphology |
| CN101059836A (en) * | 2007-06-01 | 2007-10-24 | 华南理工大学 | Human eye positioning and human eye state recognition method |
| CN101901472A (en) * | 2010-07-07 | 2010-12-01 | 清华大学 | Method for aligning non-rigid robust batch images based on matrix rank minimization |
| CN101923645A (en) * | 2009-06-09 | 2010-12-22 | 黑龙江大学 | Iris splitting method suitable for low-quality iris image in complex application context |
| CN102332098A (en) * | 2011-06-15 | 2012-01-25 | 夏东 | Method for pre-processing iris image |
-
2013
- 2013-11-15 CN CN201310570892.4A patent/CN103632137B/en not_active Expired - Fee Related
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20060280361A1 (en) * | 2005-06-14 | 2006-12-14 | Canon Kabushiki Kaisha | Image processing apparatus, image processing method,computer program, and storage medium |
| CN1900951A (en) * | 2006-06-02 | 2007-01-24 | 哈尔滨工业大学 | Iris image flexible specification method based on mathematical morphology |
| CN101059836A (en) * | 2007-06-01 | 2007-10-24 | 华南理工大学 | Human eye positioning and human eye state recognition method |
| CN101923645A (en) * | 2009-06-09 | 2010-12-22 | 黑龙江大学 | Iris splitting method suitable for low-quality iris image in complex application context |
| CN101901472A (en) * | 2010-07-07 | 2010-12-01 | 清华大学 | Method for aligning non-rigid robust batch images based on matrix rank minimization |
| CN102332098A (en) * | 2011-06-15 | 2012-01-25 | 夏东 | Method for pre-processing iris image |
Non-Patent Citations (1)
| Title |
|---|
| 徐锋 等: "基于肤色检测和人眼定位的人脸检测方法", 《计算机***应用》 * |
Cited By (16)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN104573660A (en) * | 2015-01-13 | 2015-04-29 | 青岛大学 | Method for precisely positioning human eyes by SIFT point descriptor |
| AU2017337317B2 (en) * | 2016-09-29 | 2022-12-08 | Magic Leap, Inc. | Neural network for eye image segmentation and image quality estimation |
| CN110022753B (en) * | 2016-09-29 | 2022-02-25 | 奇跃公司 | Neural network for eye image segmentation and image quality estimation |
| CN110022753A (en) * | 2016-09-29 | 2019-07-16 | 奇跃公司 | Neural network for eye image segmentation and picture quality estimation |
| US11776131B2 (en) | 2016-09-29 | 2023-10-03 | Magic Leap, Inc. | Neural network for eye image segmentation and image quality estimation |
| WO2019104616A1 (en) * | 2017-11-30 | 2019-06-06 | 深圳配天智能技术研究院有限公司 | Image processing method and device and computer-readable storage medium |
| CN109165306B (en) * | 2018-08-09 | 2021-11-23 | 长沙理工大学 | Image retrieval method based on multitask Hash learning |
| CN109165306A (en) * | 2018-08-09 | 2019-01-08 | 长沙理工大学 | Image search method based on the study of multitask Hash |
| CN109376649A (en) * | 2018-10-20 | 2019-02-22 | 张彦龙 | A method of likelihood figure, which is reduced, from eye gray level image calculates the upper lower eyelid of identification |
| CN110619273A (en) * | 2019-08-14 | 2019-12-27 | 中山市奥珀金属制品有限公司 | Efficient iris recognition method and device |
| CN110619273B (en) * | 2019-08-14 | 2023-10-31 | 张杰辉 | Efficient iris recognition method and recognition device |
| CN110781745B (en) * | 2019-09-23 | 2022-02-11 | 杭州电子科技大学 | Tail eyelash detection method based on composite window and gradient weighted direction filtering |
| CN110781745A (en) * | 2019-09-23 | 2020-02-11 | 杭州电子科技大学 | Tail eyelash detection method based on composite window and gradient weighted direction filtering |
| CN111144413A (en) * | 2019-12-30 | 2020-05-12 | 福建天晴数码有限公司 | Iris positioning method and computer readable storage medium |
| CN111241951A (en) * | 2020-01-03 | 2020-06-05 | 中山市奥珀金属制品有限公司 | Iris image processing method and device |
| CN111241951B (en) * | 2020-01-03 | 2023-10-31 | 张杰辉 | Iris image processing method and device |
Also Published As
| Publication number | Publication date |
|---|---|
| CN103632137B (en) | 2016-08-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN103632137A (en) | Human iris image segmentation method | |
| CN106446952B (en) | A kind of musical score image recognition methods and device | |
| Hauagge et al. | Image matching using local symmetry features | |
| CN102360421B (en) | Face identification method and system based on video streaming | |
| CN102902967B (en) | Method for positioning iris and pupil based on eye structure classification | |
| CN106156712A (en) | A kind of based on the ID (identity number) card No. recognition methods under natural scene and device | |
| CN107092871B (en) | Remote sensing image building detection method based on multiple dimensioned multiple features fusion | |
| CN110443128B (en) | Finger vein identification method based on SURF feature point accurate matching | |
| CN103093215A (en) | Eye location method and device | |
| CN104376548A (en) | Fast image splicing method based on improved SURF algorithm | |
| CN103971126A (en) | Method and device for identifying traffic signs | |
| CN103440476A (en) | Locating method for pupil in face video | |
| CN104217221A (en) | Method for detecting calligraphy and paintings based on textural features | |
| CN104268853A (en) | Infrared image and visible image registering method | |
| CN109978848A (en) | Method based on hard exudate in multiple light courcess color constancy model inspection eye fundus image | |
| CN108520214A (en) | A kind of finger vein identification method based on multiple dimensioned HOG and SVM | |
| CN104680130A (en) | Chinese character recognition method for identification cards | |
| CN103218605A (en) | Quick eye locating method based on integral projection and edge detection | |
| CN104268598A (en) | Human leg detection method based on two-dimensional scanning lasers | |
| CN113392856B (en) | Image forgery detection device and method | |
| CN106709500A (en) | Image feature matching method | |
| CN105225216A (en) | Based on the Iris preprocessing algorithm of space apart from circle mark rim detection | |
| CN112733711B (en) | Remote sensing image damaged building extraction method based on multi-scale scene change detection | |
| CN103198319A (en) | Method of extraction of corner of blurred image in mine shaft environment | |
| Li | An Iris recognition algorithm based on coarse and fine location |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | 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: 20160824 Termination date: 20171115 |
|
| CF01 | Termination of patent right due to non-payment of annual fee |