CN111743510B - Human eye Hartmann facula image denoising method based on clustering - Google Patents
Human eye Hartmann facula image denoising method based on clustering Download PDFInfo
- Publication number
- CN111743510B CN111743510B CN202010584274.5A CN202010584274A CN111743510B CN 111743510 B CN111743510 B CN 111743510B CN 202010584274 A CN202010584274 A CN 202010584274A CN 111743510 B CN111743510 B CN 111743510B
- Authority
- CN
- China
- Prior art keywords
- image
- human eye
- clustering
- hartmann
- matrix
- 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.)
- Active
Links
Classifications
-
- A—HUMAN NECESSITIES
- A61—MEDICAL OR VETERINARY SCIENCE; HYGIENE
- A61B—DIAGNOSIS; SURGERY; IDENTIFICATION
- A61B3/00—Apparatus for testing the eyes; Instruments for examining the eyes
- A61B3/10—Objective types, i.e. instruments for examining the eyes independent of the patients' perceptions or reactions
- A61B3/14—Arrangements specially adapted for eye photography
Abstract
The invention provides a clustering-based human eye Hartmann facula image denoising method, which takes a human eye Hartmann facula image as input, firstly, performs classification clustering on image data, and divides the image data into background and foreground types, thereby adaptively removing image background noise and camera CCD readout noise; secondly, respectively calculating the number of all data points belonging to the foreground class in the row and the column of each data point aiming at the foreground class data, and constructing a row and column counting matrix; and finally, respectively performing two-class clustering on the row and column counting matrixes, and dividing data into normal human eye aberration light spots, optical devices and cornea reflection light spots, so that the optical devices and human eye cornea reflection noise are adaptively removed, and the drying treatment of the whole image is completed. The invention uses the image processing method to remove the image noise of the human eye Hartmann facula, thereby improving the detection precision of human eye aberration; the complexity and cost of the system are reduced, and meanwhile, the method is easy to use without manually setting parameters.
Description
Technical Field
The invention relates to the application field of a shack-Hartmann wavefront sensing measurement technology in human eye adaptive optics, in particular to a clustering-based human eye Hartmann facula image denoising method which can eliminate the influence of background noise, camera readout noise, optical device reflection noise and human eye cornea reflection noise on human eye aberration measurement.
Background
The human eye aberration contains not only low-order aberration such as defocus and astigmatism, but also non-negligible high-order aberration components. The self-adaptive optical technology can simultaneously realize the accurate measurement and correction of the lower-order and higher-order aberrations of the human eyes, and directly promotes the development of fundus imaging and vision optics. In the existing human eye self-adaptive optical system, a Hartmann wavefront sensor is widely adopted to acquire a human eye aberration facula image, and a wavefront restoration technology is used for measuring human eye aberration based on the image, so that a wavefront corrector is driven to compensate and correct the aberration, and the influence of the human eye aberration on the whole system is eliminated. However, since various noises exist in the human eye adaptive optics system such as: camera readout noise, random background noise, system optics reflection noise, human eye cornea reflection noise, etc., result in the obtained original human eye aberration hartmann spot image being undesirable. Therefore, how to remove various noises in the Hartmann light spot image, so that the measurement accuracy of the system to human eye aberration is particularly important to the whole human eye self-adaptive optical system.
The Hartmann facula image is simultaneously affected by various noises, wherein random noises and background noises can be easily removed by adopting threshold value de-dryness, filtering de-dryness and the like, and reflection of light of an optical device and reflection of light of cornea are difficult to eliminate by adopting a common threshold value method. Conventional solutions typically adjust the hardware of an optical system or specially design the entire optical system, such as adding annular illumination, using a reflective optical system instead of a transmissive system, making the incident light incident off-center, etc., however, these methods increase the complexity and cost of the system, making the system difficult to adjust, and even impossible to implement in some specific situations. Aiming at the problem that various reflection noise in human eye Hartmann light spots is difficult to eliminate, the invention provides a clustering-based human eye Hartmann light spot image drying method, which uses an image processing technology to remove various noises in a human eye Hartmann light spot image, and improves the detection precision of human eye aberration; compared with the traditional method based on hardware transformation and special customization of the optical system, the method reduces the complexity and cost of the system, and meanwhile, the method does not need to set parameters manually and is easy to use.
Disclosure of Invention
The invention solves the technical problems that: the method for denoising the human eye Hartmann facula image based on clustering is provided for overcoming the defects of the prior art, and various noises in the human eye Hartmann facula image are removed, so that the detection precision of human eye aberration is improved.
The technical scheme adopted for solving the technical problems is as follows: a human eye Hartmann facula image denoising method based on clustering comprises the following steps:
step (1), acquiring an original image I (x, y) of a human eye Hartmann facula from an optical system by using a camera, wherein (x, y) is a coordinate position, and the image size is M multiplied by N;
step (2), carrying out classification clustering on the original image to obtain a classification mark matrix L (x, y), wherein L (x, y) =1 represents background noise class, and L (x, y) =2 represents foreground class;
step (3), processing the original image by using a marking matrix L to remove image background noise and CCD readout noise and obtain an image M (x, y); the specific method comprises the following steps: initializing M (x, y) =i (x, y), and then processing according to the following formula:
I(x,y)＝0,if L(x,y)＝1
step (4), processing the marking matrix L obtained in the step (2) to construct a row counting matrix R (x, y); the concrete construction method comprises the following steps:
step (5), performing a binary cluster on the row counting matrix R to obtain a cluster mark RL (x, y), wherein RL (x, y) =1 represents the position of the reflective optical element or the reflective spot of the cornea of the human eye, and RL (x, y) =2 represents the position of the aberration spot of the normal human eye;
step (6), processing the marking matrix L obtained in the step (2) to construct a column counting matrix C (x, y); the concrete construction method comprises the following steps:
step (7), performing classification clustering on the column counting matrix C to obtain a clustering mark CL (x, y), wherein CL (x, y) =1 represents the position of an optical device reflection or human eye cornea reflection point, and CL (x, y) =2 represents the position of a normal human eye aberration light spot;
and (8) processing the image M by using the marking matrixes RL and CL to obtain a final image F (x, y) after removing reflection noise of an optical device and reflection noise of cornea of human eyes, wherein the specific method comprises the following steps of: initializing F (x, y) =m (x, y), and then processing according to the following formula:
F(x,y)＝0,if RL(x,y)＝1||CL(x,y)＝1
compared with the prior art, the invention has the advantages that:
(1) According to the invention, no additional hardware modification and optical path design are required to be carried out on the system, and the cost and complexity of the system are not increased due to pure software calculation;
(2) The invention is fully and automatically carried out, does not need to manually set various parameters, and is easy to use;
(3) The invention has strong adaptability and can be used on any Hartmann facula image with background noise and reflection noise.
Drawings
FIG. 1 is a flow chart of a clustering-based human eye Hartmann spot image denoising method;
fig. 2 shows images before and after the de-drying process for a specific human eye hartmann spot image according to the present invention.
Detailed Description
In order to clearly illustrate the implementation of the invention in detail, specific embodiments of the invention are given below. The following detailed description of preferred embodiments of the invention, with reference to the accompanying drawings, omits details and functions that are not necessary for the invention in the description so as to prevent confusion of the understanding of the invention.
The flow chart of the clustering-based human eye Hartmann facula image denoising method is shown in fig. 1, and the following descriptions are provided:
according to the method, an original human eye Hartmann facula image is obtained through system hardware, as shown in fig. 2 (a), and the original image is influenced by various background noise and reflective noise at the same time, and the wavefront restoration can influence the measurement accuracy of human eye aberration by directly using the image. The algorithm will perform a de-drying process on the image by:
step (1), carrying out classification clustering on an original image to obtain a classification mark matrix L (x, y), wherein L (x, y) =1 represents background noise class, and L (x, y) =2 represents foreground class;
step (2), processing an original image by using a marking matrix L to remove image background noise and CCD readout noise and obtain an image M (x, y); the specific method comprises the following steps: initializing M (x, y) =i (x, y), and then processing according to the following formula:
I(x,y)＝0,if L(x,y)＝1
the result of obtaining an image after removing the background noise and the CCD readout noise after this step is shown in fig. 2 (b).
Step (3), processing the marking matrix L obtained in the step (1) to construct a row counting matrix R (x, y); the concrete construction method comprises the following steps:
step (4), performing classification clustering on the row counting matrix R to obtain a clustering mark RL (x, y), wherein RL (x, y) =1 represents the position of an optical device reflection or human eye cornea reflection point, and RL (x, y) =2 represents the position of a normal human eye aberration light spot;
step (5), processing the marking matrix L obtained in the step (1) to construct a column counting matrix C (x, y); the concrete construction method comprises the following steps:
step (6), performing classification clustering on the column counting matrix C to obtain a clustering mark CL (x, y), wherein CL (x, y) =1 represents the position of an optical device reflection or human eye cornea reflection point, and CL (x, y) =2 represents the position of a normal human eye aberration light spot;
and (7) processing the image M by using the marking matrixes RL and CL to obtain a final image F (x, y) after removing reflection noise of an optical device and reflection noise of cornea of human eyes, wherein the specific method comprises the following steps of: initializing F (x, y) =m (x, y), and then processing according to the following formula:
F(x,y)＝0,if RL(x,y)＝1||CL(x,y)＝1
after the processing of the above steps, as shown in fig. 2 (c), the final image F (x, y) is shown, and it can be seen that the background noise and the reflection noise in the original image have been effectively removed.
The invention has been described with reference to the preferred embodiments. It should be understood that various other changes, substitutions, and alterations can be made by those skilled in the art without departing from the spirit and scope of the invention. Accordingly, the scope of the invention is not limited to the specific embodiments described above, but should be defined by the appended claims.
Claims (3)
1. A human eye Hartmann facula image denoising method based on clustering is characterized in that: the method comprises the following steps:
step (1), acquiring an original image I (x, y) of a human eye Hartmann facula from an optical system by using a camera, wherein (x, y) is a coordinate position, and the image size is M multiplied by N;
step (2), carrying out classification clustering on the original image to obtain a classification mark matrix L (x, y), wherein L (x, y) =1 represents background noise class, and L (x, y) =2 represents foreground class;
step (3), processing the original image by using a marking matrix L to remove image background noise and CCD readout noise and obtain an image M (x, y);
step (4), processing the marking matrix L obtained in the step (2) to construct a row counting matrix R (x, y);
step (5), performing a binary cluster on the row counting matrix R to obtain a cluster mark RL (x, y), wherein RL (x, y) =1 represents the position of the reflective optical element or the reflective spot of the cornea of the human eye, and RL (x, y) =2 represents the position of the aberration spot of the normal human eye;
step (6), processing the marking matrix L obtained in the step (2) to construct a column counting matrix C (x, y);
step (7), performing classification clustering on the column counting matrix C to obtain a clustering mark CL (x, y), wherein CL (x, y) =1 represents the position of an optical device reflection or human eye cornea reflection point, and CL (x, y) =2 represents the position of a normal human eye aberration light spot;
and (8) processing the image M by using the cluster mark RL and the cluster mark CL to obtain a final image F (x, y) after removing reflection noise of an optical device and reflection noise of cornea of human eyes, wherein the specific method for obtaining the final image F (x, y) by using the mark matrixes RL, CL and the image M comprises the following steps: initializing F (x, y) =m (x, y), and then processing according to the following formula:
F(x,y)＝0,ifRL(x,y)＝1||CL(x,y)＝1。
2. the clustering-based human eye Hartmann spot image denoising method as set forth in claim 1, wherein: the specific method for obtaining the image M through the marking matrix L and the original image I in the step (3) is as follows: initializing M (x, y) =i (x, y), and then processing according to the following formula:
I(x,y)＝0,ifL(x,y)＝1。
3. the clustering-based human eye Hartmann spot image denoising method as set forth in claim 1, wherein: the specific method for constructing the row counting matrix R (x, y) and the column counting matrix C (x, y) in the step (4) and the step (6) is as follows:
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010584274.5A CN111743510B (en) | 2020-06-24 | 2020-06-24 | Human eye Hartmann facula image denoising method based on clustering |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202010584274.5A CN111743510B (en) | 2020-06-24 | 2020-06-24 | Human eye Hartmann facula image denoising method based on clustering |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN111743510A CN111743510A (en) | 2020-10-09 |
| CN111743510B true CN111743510B (en) | 2023-09-19 |
Family
ID=72676999
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202010584274.5A Active CN111743510B (en) | 2020-06-24 | 2020-06-24 | Human eye Hartmann facula image denoising method based on clustering |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN111743510B (en) |
Families Citing this family (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114708264B (en) * | 2022-06-06 | 2022-08-12 | 广东工业大学 | Light spot quality judging method, device, equipment and storage medium |
Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103278247A (en) * | 2013-05-06 | 2013-09-04 | 中国科学院光电技术研究所 | Background noise characteristic estimation method for point source Hartmann wavefront detector |
| CN103745257A (en) * | 2013-12-23 | 2014-04-23 | 温州医科大学眼视光器械有限公司 | Cone-cell density calculation method based on image identification |
| FR3006087A1 (en) * | 2013-05-21 | 2014-11-28 | Miratech | PROCESSING DIGITAL DATA REPRESENTATIVE OF THE OCULAR BEHAVIOR OF A SET OF USERS |
| CN105283118A (en) * | 2013-06-10 | 2016-01-27 | 埃西勒国际通用光学公司 | Method for determining wave-front aberration data of a to-be-tested optical system |
| CN106097255A (en) * | 2016-05-26 | 2016-11-09 | 中国科学院光电技术研究所 | Background noise feature assessment method for point source Hartman wavefront detector |
| CN107478174A (en) * | 2017-07-12 | 2017-12-15 | 江南大学 | A kind of Shack Hartmann sensor centroid detection method for dark weak signal |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2003022138A1 (en) * | 2001-09-07 | 2003-03-20 | Kabushiki Kaisha Topcon | Eye optical characteristic measuring instrument |
| JP6679505B2 (en) * | 2014-06-16 | 2020-04-15 | エージェンシー フォー サイエンス，テクノロジー アンド リサーチ | Reduction of speckle noise in optical coherence tomography images |
| DE112016000356T5 (en) * | 2015-01-16 | 2018-01-11 | Imra Europe S.A.S. | Self-rectification of stereo cameras |
-
2020
- 2020-06-24 CN CN202010584274.5A patent/CN111743510B/en active Active
Patent Citations (6)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103278247A (en) * | 2013-05-06 | 2013-09-04 | 中国科学院光电技术研究所 | Background noise characteristic estimation method for point source Hartmann wavefront detector |
| FR3006087A1 (en) * | 2013-05-21 | 2014-11-28 | Miratech | PROCESSING DIGITAL DATA REPRESENTATIVE OF THE OCULAR BEHAVIOR OF A SET OF USERS |
| CN105283118A (en) * | 2013-06-10 | 2016-01-27 | 埃西勒国际通用光学公司 | Method for determining wave-front aberration data of a to-be-tested optical system |
| CN103745257A (en) * | 2013-12-23 | 2014-04-23 | 温州医科大学眼视光器械有限公司 | Cone-cell density calculation method based on image identification |
| CN106097255A (en) * | 2016-05-26 | 2016-11-09 | 中国科学院光电技术研究所 | Background noise feature assessment method for point source Hartman wavefront detector |
| CN107478174A (en) * | 2017-07-12 | 2017-12-15 | 江南大学 | A kind of Shack Hartmann sensor centroid detection method for dark weak signal |
Non-Patent Citations (2)
| Title |
|---|
| 兰善伟.基于瞳孔角膜反射视线追踪技术的研究.中国优秀硕士学位论文全文数据库.2016,全文. * |
| 段海峰 等.模式正交性对哈特曼-夏克传感器波前测量等的影响.光学学报.2003,第23卷(第9期),全文. * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN111743510A (en) | 2020-10-09 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Abdelhamed et al. | A high-quality denoising dataset for smartphone cameras | |
| RU2716843C1 (en) | Digital correction of optical system aberrations | |
| KR101610975B1 (en) | Single-lens extended depth-of-field imaging systems | |
| US11195257B2 (en) | Image processing method, image processing apparatus, imaging apparatus, lens apparatus, storage medium, and image processing system | |
| US8594388B2 (en) | Large depth-of-field imaging system and iris recogniton system | |
| WO2009141998A1 (en) | Calibration method, calibration device, and calibration system having the device | |
| US20180165501A1 (en) | Extended depth-of-field biometric system | |
| EP2228677A1 (en) | Extended depth-of-field surveillance imaging system | |
| US8416334B2 (en) | Thick single-lens extended depth-of-field imaging systems | |
| JP2008245157A (en) | Imaging device and method therefor | |
| CN102082912A (en) | Image capturing apparatus and image processing method | |
| US10778915B2 (en) | Dual-aperture ranging system | |
| CN111743510B (en) | Human eye Hartmann facula image denoising method based on clustering | |
| JPWO2012086127A1 (en) | Image processing apparatus, imaging apparatus, and image processing method | |
| KR20160061690A (en) | Apparatus and Method for Detecting Eyes | |
| CN115456914A (en) | Priori knowledge-based defocused image deblurring method, device and medium | |
| US20180338049A1 (en) | Image processing apparatus performing image recovery processing, imaging apparatus, image processing method, and storage medium | |
| CN111222446B (en) | Face recognition method, face recognition device and mobile terminal | |
| CN110827375B (en) | Infrared image true color coloring method and system based on low-light-level image | |
| JP7191588B2 (en) | Image processing method, image processing device, imaging device, lens device, program, and storage medium | |
| JP4818956B2 (en) | Imaging apparatus and method thereof | |
| JP7254440B2 (en) | Image processing device, imaging device, image processing method, and program | |
| Lluis-Gomez et al. | Chromatic aberration correction in RAW domain for image quality enhancement in image sensor processors | |
| CN113489907B (en) | Imaging control method and device of thermal infrared imager and infrared imaging equipment | |
| CN110728714B (en) | Image processing method and device, storage medium and electronic equipment |
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 |