CN109255753A - A kind of eye fundus image joining method - Google Patents
A kind of eye fundus image joining method Download PDFInfo
- Publication number
- CN109255753A CN109255753A CN201810980643.5A CN201810980643A CN109255753A CN 109255753 A CN109255753 A CN 109255753A CN 201810980643 A CN201810980643 A CN 201810980643A CN 109255753 A CN109255753 A CN 109255753A
- Authority
- CN
- China
- Prior art keywords
- image
- eye fundus
- fundus image
- characteristic point
- point
- 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
- 238000000034 method Methods 0.000 title claims abstract description 33
- 239000000284 extract Substances 0.000 claims abstract description 8
- 230000008569 process Effects 0.000 claims abstract description 7
- 238000007781 pre-processing Methods 0.000 claims abstract description 5
- 230000004927 fusion Effects 0.000 claims description 20
- 238000001914 filtration Methods 0.000 claims description 17
- 238000005457 optimization Methods 0.000 claims description 7
- 238000003672 processing method Methods 0.000 claims description 6
- 230000002708 enhancing effect Effects 0.000 claims description 4
- 230000003313 weakening effect Effects 0.000 claims description 3
- 230000000694 effects Effects 0.000 abstract description 6
- 230000036541 health Effects 0.000 abstract description 6
- 238000010586 diagram Methods 0.000 description 11
- 238000003745 diagnosis Methods 0.000 description 2
- 230000004048 modification Effects 0.000 description 2
- 238000012986 modification Methods 0.000 description 2
- 238000012014 optical coherence tomography Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 239000008280 blood Substances 0.000 description 1
- 210000004369 blood Anatomy 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 230000008859 change Effects 0.000 description 1
- 230000010485 coping Effects 0.000 description 1
- 238000005516 engineering process Methods 0.000 description 1
- 238000000605 extraction Methods 0.000 description 1
- 208000030533 eye disease Diseases 0.000 description 1
- 238000005286 illumination Methods 0.000 description 1
- 238000003384 imaging method Methods 0.000 description 1
- 238000002844 melting Methods 0.000 description 1
- 230000008018 melting Effects 0.000 description 1
- 238000002156 mixing Methods 0.000 description 1
- 230000001225 therapeutic effect Effects 0.000 description 1
- 230000007704 transition Effects 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T3/00—Geometric image transformations in the plane of the image
- G06T3/40—Scaling of whole images or parts thereof, e.g. expanding or contracting
- G06T3/4038—Image mosaicing, e.g. composing plane images from plane sub-images
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/10—Image acquisition modality
- G06T2207/10004—Still image; Photographic image
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/20—Special algorithmic details
- G06T2207/20024—Filtering details
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06T—IMAGE DATA PROCESSING OR GENERATION, IN GENERAL
- G06T2207/00—Indexing scheme for image analysis or image enhancement
- G06T2207/30—Subject of image; Context of image processing
- G06T2207/30004—Biomedical image processing
- G06T2207/30041—Eye; Retina; Ophthalmic
Landscapes
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Engineering & Computer Science (AREA)
- Theoretical Computer Science (AREA)
- Image Processing (AREA)
- Eye Examination Apparatus (AREA)
Abstract
The invention discloses a kind of eye fundus image joining methods, comprising the following steps: S1 shoots one group of eye fundus image；S2 pre-processes the eye fundus image of shooting；S3 extracts the characteristic point after pre-processing in eye fundus image；S4 finds the corresponding relationship of characteristic point between edge graph and Centered Graphs according to the preset fixed figure of user；S5 aligns failure if pairs of characteristic point quantity is not more than 2 pairs, terminates picture mosaic；Otherwise offset mean value Δ x and Δ y of the characteristic point relative to the characteristic point in Centered Graphs in edge graph are calculated, and according to offset mean value Δ x and Δ y combination of edge figure and Centered Graphs.The present invention is able to ascend field range, increases effective information, greatly enhances the effect of diagnosing and treating.And it reduces multi-angle repeatedly to shoot on a large scale, greatly reduces the quality time of health care workers, improve the efficiency of curative activity.
Description
Technical field
The present invention relates to a kind of eye fundus image processing technology fields, more particularly to a kind of eye fundus image joining method.
Background technique
Nowadays, as health is increasingly taken seriously, requirement of the people to medical and health conditions is higher and higher, and medical treatment is single
Position and demand of the health care workers to medical instrument are increasing, it is desirable that also higher and higher, field range is limited to be diagnosed and control
Treat the more and more not competent high-precision of equipment, high-frequency work.
But at present in eye disease diagnosis and therapeutic equipment, optical coherence tomography scanner (abbreviation OCT), crack
The equipment such as lamp, Ultrasonic pachymetry number instrument, confocal laser eyeground angiographic instrument, fundus camera, eyeground blood sample Concentration Testing are all
Small in the prevalence of field range, the few problem of effective information greatly affected the effect of diagnosis or treatment.
Such issues that solve the best way is exactly the field range for increasing equipment imaging, but this usually requires to spend
Higher expense or longer R&D cycle.So the main method for coping with such problem at present is that multi-angle is a wide range of
Ground is repeatedly shot, but is observed these images again and can greatly be expended the quality time of health care workers, and curative activity is seriously reduced
Efficiency.
Summary of the invention
The present invention is directed at least solve the technical problems existing in the prior art, a kind of eyeground is especially innovatively proposed
Image split-joint method.
In order to realize above-mentioned purpose of the invention, the invention discloses a kind of eye fundus image joining methods, including following step
It is rapid:
S1 shoots one group of eye fundus image；
S2 pre-processes the eye fundus image of shooting；
S3 extracts the characteristic point after pre-processing in eye fundus image；
S4 finds the corresponding relationship of characteristic point between edge graph and Centered Graphs according to the preset fixed figure of user；
S5 aligns failure if pairs of characteristic point quantity is not more than 2 pairs, terminates picture mosaic；Otherwise it calculates in edge graph
Offset mean value Δ x and Δ y of the characteristic point relative to the characteristic point in Centered Graphs, and according to offset mean value Δ x and Δ y
Combination of edge figure and Centered Graphs.
In the preferred embodiment of the present invention, the preprocess method of the eye fundus image in step S2 includes following
One of or any combination sequence are as follows:
S21 carries out image gray processing processing to eye fundus image；
S22 carries out enhancing contrast processing to eye fundus image；
S23 carries out weakening noise processed to eye fundus image.
In the preferred embodiment of the present invention, enhancing contrast processing is carried out to eye fundus image in step S22
Method are as follows: utilize histogram equalization or limitation contrast self-adapting histogram equilibrium algorithm；
Or/and the method for weaken to eye fundus image noise processed in step S23 are as follows: utilize gaussian filtering or mean value
Filtering algorithm.
In the preferred embodiment of the present invention, the characteristic point after pre-processing in eye fundus image is extracted in step S3
The following steps are included:
S31 calculates image in the gradient in the direction x and y, obtains x direction gradient image lx and y direction gradient image ly；
S32 calculates and optimizes a square processing to x direction gradient image lx, obtains lx2=lx2, to y direction gradient figure
As ly optimizes square processing, ly2=ly is obtained2, it merges the direction x and y direction gradient image optimizes gradient processing,
Obtain lxy=lx*ly；
S33 calculates single channel imageWherein lxy2=lxy2 is to merge the direction x and the direction y
Square processing image；
S34, will be greater than the pixel of threshold value t as characteristic point in single channel image h, wherein threshold value 0≤t≤255.This
The feature point extraction algorithm of invention has illumination invariant, rotational invariance, scale invariability, significantly reduces omission factor,
Improve robustness.And Feature Points Matching algorithm steps are simple, and calculation amount is minimum, significantly reduce operation time, improve
Processing speed.
In the preferred embodiment of the present invention, step S33 are as follows:
Image lx2, ly2 and lxy after optimization processing is filtered, obtain x trend pass filtering processing image gx2,
Y trend pass filtering handles image gy2 and the fusion direction x and y trend pass filtering handles image gxy；
Calculate single channel imageWherein gxy2=gxy2 is to merge the direction x and the filter of the direction y
Popin side handles image.It is filtered that be conducive to image smoother to the image after optimization processing.
In the preferred embodiment of the present invention, image lx2, ly2 and lxy after optimization processing are filtered
Processing is gaussian filtering.
In the preferred embodiment of the present invention, characteristic point between edge graph and Centered Graphs is found in step S4
The processing method of corresponding relationship the following steps are included:
S41, centered on characteristic point s, with k1There is k on boundary for the circle of radius2A pixel, wherein characteristic point s is one
Any one characteristic point in width eye fundus image；The k1For positive number, k2For positive integer；
S42 is labeled as 1 if the pixel value of the boundary point of circle is greater than the pixel value of characteristic point s, otherwise, is labeled as 0；
S43, if some characteristic point in edge graph has continuous n same position with some characteristic point in Centered Graphs
Mark identical, then the two characteristic points are a pair of of match point, wherein the n is no more than k2Positive integer, preferably n=9.
In the preferred embodiment of the present invention, the k1It is 3, k2It is 16.
In the preferred embodiment of the present invention, in step S5 combination of edge figure and Centered Graphs processing method packet
Include following steps:
S51, establish it is one wide centered on scheme width k3Again, the high k of a height of Centered Graphs4Null images again are described as fusion figure
k3、k4For the positive number not less than 1 and at least one is not less than 2, and Centered Graphs are copied in fusion figure；
S52 in edge graph from top to bottom from left to right, successively traverses pixel, calculates point (x, y) and adds offset
Mean value Δ x and Δ y to Centered Graphs centre coordinate (x0,y0) distance
S53, judges whether d1 is greater than Centered Graphs radius r, that is, judge the point (x, y) whether in the range of Centered Graphs, if
D1 is greater than r, i.e. the pixel value c2 of the point (x, y) is then directly copied to the phase in fusion figure outside Centered Graphs by the point (x, y)
Answer position；Otherwise, the distance that the point (x, y) arrives edge graph center is calculated
Wherein, (x0′,y0 ′) it is edge graph centre coordinate, then the point (x, y) is melting
Pixel value in conjunction figure isWherein, c1 is the point (x, y) at Centered Graphs center
Coordinate (x0,y0) at pixel value, c2 be coordinate (pixel value at x+ Δ x, y+ Δ y) place, p of the point (x, y) in edge graph
For positive integer, p as an adjustment factor, during the bigger image co-registration of value feature cross get over it is smooth.Image of the invention
Blending algorithm does not have obvious splicing boundary, and transition is naturally smooth.
In the preferred embodiment of the present invention, the k3=k4=3；
Or step S51 be establish it is one wide centered on scheme width k3Again, the high k of a height of Centered Graphs4Null images again are as fusion
Figure, the k3、k4For the positive integer not less than 3, Centered Graphs are copied into fusion figure centre.
In conclusion by adopting the above-described technical solution, the beneficial effects of the present invention are: the present invention is able to ascend view
Wild range increases effective information, greatly enhances the effect of diagnosing and treating.And it reduces multi-angle repeatedly to clap on a large scale
It takes the photograph, greatly reduces the quality time of health care workers, improve the efficiency of curative activity.
Detailed description of the invention
Fig. 1 is flow diagram of the present invention.
(a)~(i) is one group of eye fundus image schematic diagram that the present invention is shot in Fig. 2.
Fig. 3 is the present invention through image gray processing treated image schematic diagram.
Fig. 4 is the enhanced contrast of the present invention treated image schematic diagram.
Fig. 5 is image schematic diagram of the present invention after weakening noise processed.
Fig. 6 is the image schematic diagram that the present invention extracts edge graph characteristic point.
Fig. 7 is the image schematic diagram that the present invention extracts Centered Graphs characteristic point.
Fig. 8 is the correspondence diagram of characteristic point between edge graph and Centered Graphs of the present invention.
Fig. 9 is the present invention centered on characteristic point s, to there is 16 pixel schematic diagrames on 3 boundary for the circle of radius.
Figure 10 is that the pixel value judging result of the pixel value of the boundary point of circle and characteristic point label is illustrated in Fig. 9 of the present invention
Figure.
Figure 11 is that edge graph of the present invention with Centered Graphs merges schematic diagram.
Figure 12 is that image co-registration of the present invention completes schematic diagram.
Specific embodiment
The embodiment of the present invention is described below in detail, examples of the embodiments are shown in the accompanying drawings, wherein from beginning to end
Same or similar label indicates same or similar element or element with the same or similar functions.Below with reference to
The embodiment of attached drawing description is exemplary, and for explaining only the invention, and is not considered as limiting the invention.
The invention discloses a kind of eye fundus image joining methods, as shown in Figure 1, comprising the following steps:
The first step shoots one group of eye fundus image, as shown in Fig. 2, this group of eye fundus image is the same of same fundus camera shooting
The continuous nine width eye fundus image on one eyeground.
Second step pre-processes the eye fundus image of shooting.In the present embodiment, successively to eye fundus image image
Gray processing, as shown in Figure 3；And pass through histogram equalization (or limitation contrast self-adapting histogram equilibrium scheduling algorithm) enhancing
Contrast, as shown in Figure 4；Noise is weakened by gaussian filtering (or mean filter scheduling algorithm) again, as shown in Figure 5.
Third step extracts the characteristic point after pre-processing in eye fundus image, as shown in Figures 6 and 7.
In the present embodiment, extract the characteristic point after pretreatment in eye fundus image specifically includes the following steps:
S31 calculates image in the gradient in the direction x and y, obtains x direction gradient image lx and y direction gradient image ly；
S32 calculates and optimizes a square processing to x direction gradient image lx, obtains lx2=lx2, to y direction gradient figure
As ly optimizes square processing, ly2=ly is obtained2, it merges the direction x and y direction gradient image optimizes gradient processing,
Obtain lxy=lx*ly；
S33 carries out gaussian filtering process to image lx2, ly2 and lxy after optimization processing, obtains the direction x gaussian filtering
Handle the direction image gx2, y gaussian filtering process image gy2 and the fusion direction x and the direction y gaussian filtering process image gxy；
Calculate single channel imageWherein gxy2=gxy2 is to merge the direction x and the direction y height
This filtering square processing image；
S34, will be greater than the pixel of threshold value t as characteristic point in single channel image h, wherein threshold value 0≤t≤255.It is excellent
Selecting pixel threshold value t is 100 or 200.
4th step finds the corresponding pass of characteristic point between edge graph and Centered Graphs according to the preset fixed figure of user
System, as shown in Figure 8.
In this hair embodiment, the processing method for finding the corresponding relationship of characteristic point between edge graph and Centered Graphs has
Body the following steps are included:
S41, centered on characteristic point s, with k1There is k on boundary for the circle of radius2A pixel, wherein characteristic point s is one
Any one characteristic point in width eye fundus image；The k1For positive number, k2For positive integer.Preferably, centered on characteristic point s, with 3
There are 16 pixels on boundary for the circle of radius, as shown in Figure 9.
S42 is labeled as 1 if the pixel value of the boundary point of circle is greater than the pixel value of characteristic point s, otherwise, is labeled as 0；
As shown in Figure 10.
S43, if some characteristic point in edge graph has continuous n same position with some characteristic point in Centered Graphs
Mark identical, then the two characteristic points are a pair of of match point, wherein the n is no more than k2Positive integer, preferably n=9.
Have the label (1) of continuous 10 same positions in Figure 10, there is the label (0) of continuous 3 same positions, have continuous 2 it is identical
The label (1) of position.
5th step aligns failure if pairs of characteristic point quantity is not more than 2 pairs, terminates picture mosaic；Otherwise edge is calculated
Offset mean value Δ x and Δ y of the characteristic point relative to the characteristic point in Centered Graphs in figure, and according to offset mean value Δ x and
Δ y combination of edge figure and Centered Graphs, as shown in figure 11.
In the present embodiment, the processing method of combination of edge figure and Centered Graphs specifically includes the following steps:
S51, establish it is one wide centered on scheme width k3Again, the high k of a height of Centered Graphs4Null images again are described as fusion figure
k3、k4For the positive number not less than 1 and at least one is not less than 2, and Centered Graphs are copied in fusion figure；Preferably, one is established
The long and wide null images for being all Fig. 3 times of center are opened as fusion figure, and Centered Graphs are copied into fusion figure centre.
S52 in edge graph from top to bottom from left to right, successively traverses pixel, calculates point (x, y) and adds offset
Mean value Δ x and Δ y to Centered Graphs centre coordinate (x0,y0) distance
S53, judges whether d1 is greater than Centered Graphs radius r, that is, judge the point (x, y) whether in the range of Centered Graphs, if
D1 is greater than r, i.e. the pixel value c2 of the point (x, y) is then directly copied to the phase in fusion figure outside Centered Graphs by the point (x, y)
Answer position；Otherwise, the distance that the point (x, y) arrives edge graph center is calculatedWherein, (x0′,
y0') it is edge graph centre coordinate, then the pixel value of the point (x, y) in fusion figure isWherein, c1 is the point (x, y) in Centered Graphs centre coordinate (x0,y0) at picture
Element value, c2 are that (pixel value at the place x+ Δ x, y+ Δ y), p is positive integer to coordinate of the point (x, y) in edge graph, and p is as one
A adjustment factor, during the bigger image co-registration of value feature cross get over it is smooth.
6th step handles remaining eye fundus image according to above-mentioned steps, as shown in figure 12.
Although an embodiment of the present invention has been shown and described, it will be understood by those skilled in the art that: not
A variety of change, modification, replacement and modification can be carried out to these embodiments in the case where being detached from the principle of the present invention and objective,
The scope of the present invention is defined by the claims and their equivalents.
Claims (10)
1. a kind of eye fundus image joining method, which comprises the following steps:
S1 shoots one group of eye fundus image；
S2 pre-processes the eye fundus image of shooting；
S3 extracts the characteristic point after pre-processing in eye fundus image；
S4 finds the corresponding relationship of characteristic point between edge graph and Centered Graphs according to the preset fixed figure of user；
S5 aligns failure if pairs of characteristic point quantity is not more than 2 pairs, terminates picture mosaic；Otherwise the feature in edge graph is calculated
Offset mean value Δ x and Δ y of the point relative to the characteristic point in Centered Graphs, and according to offset mean value Δ x and Δ y combination of edge
Figure and Centered Graphs.
2. eye fundus image joining method according to claim 1, which is characterized in that the pre- place of the eye fundus image in step S2
Reason method includes following one or any combination sequence are as follows:
S21 carries out image gray processing processing to eye fundus image；
S22 carries out enhancing contrast processing to eye fundus image；
S23 carries out weakening noise processed to eye fundus image.
3. eye fundus image joining method according to claim 2, which is characterized in that carried out in step S22 to eye fundus image
Enhance the method for contrast processing are as follows: utilize histogram equalization or limitation contrast self-adapting histogram equilibrium algorithm；
Or/and the method for weaken to eye fundus image noise processed in step S23 are as follows: calculated using gaussian filtering or mean filter
Method.
4. eye fundus image joining method according to claim 1, which is characterized in that extract eyeground after pretreatment in step S3
Characteristic point in image the following steps are included:
S31 calculates image in the gradient in the direction x and y, obtains x direction gradient image lx and y direction gradient image ly；
S32 calculates and optimizes a square processing to x direction gradient image lx, obtains lx2=lx2, to y direction gradient image ly into
Row optimization square processing, obtains ly2=ly2, merge the direction x and y direction gradient image optimize gradient processing, obtain lxy
=lx*ly；
S33 calculates single channel imageWherein lxy2=lxy2 is to merge at the direction x and the direction square y
Manage image；
S34, will be greater than the pixel of threshold value t as characteristic point in single channel image h, wherein threshold value 0≤t≤255.
5. eye fundus image joining method according to claim 4, which is characterized in that step S33 are as follows:
Image lx2, ly2 and lxy after optimization processing is filtered, the x trend pass filtering processing direction image gx2, y is obtained
Image gy2 and the fusion direction x is filtered and y trend pass filtering handles image gxy；
Calculate single channel imageWherein gxy2=gxy2 is to merge the direction x and y trend pass filtering square
Handle image.
6. eye fundus image joining method according to claim 5, which is characterized in that after optimization processing image lx2,
Ly2 and lxy is filtered as gaussian filtering.
7. eye fundus image joining method according to claim 1, which is characterized in that find edge graph and center in step S4
Between figure the corresponding relationship of characteristic point processing method the following steps are included:
S41, centered on characteristic point s, with k1There is k on boundary for the circle of radius2A pixel, wherein characteristic point s is a width eyeground
Any one characteristic point in image；The k1For positive number, k2For positive integer；
S42 is labeled as 1 if the pixel value of the boundary point of circle is greater than the pixel value of characteristic point s, otherwise, is labeled as 0；
S43, if some characteristic point in edge graph has the label phase of continuous n same position with some characteristic point in Centered Graphs
Together, then the two characteristic points are a pair of of match point, wherein the n is no more than k2Positive integer.
8. eye fundus image joining method according to claim 7, which is characterized in that the k1It is 3, k2It is 16.
9. eye fundus image joining method according to claim 1, which is characterized in that combination of edge figure and center in step S5
The processing method of figure the following steps are included:
S51, establish it is one wide centered on scheme width k3Again, the high k of a height of Centered Graphs4Null images again are as fusion figure, the k3、k4For
Positive number not less than 1 and at least one be not less than 2, Centered Graphs are copied in fusion figure；
S52 in edge graph from top to bottom from left to right, successively traverses pixel, calculates point (x, y) and adds offset mean value
Δ x and Δ y to Centered Graphs centre coordinate (x0,y0) distance
S53, judges whether d1 is greater than Centered Graphs radius r, that is, the point (x, y) is judged whether in the range of Centered Graphs, if d1 is big
In r, i.e. the pixel value c2 of the point (x, y) is then directly copied to the corresponding positions in fusion figure outside Centered Graphs by the point (x, y)
It sets；Otherwise, the distance that the point (x, y) arrives edge graph center is calculatedWherein, (x0′,y0') be
Edge graph centre coordinate, then the pixel value of the point (x, y) in fusion figure beIts
In, c1 is the point (x, y) in Centered Graphs centre coordinate (x0,y0) at pixel value, c2 be the seat of the point (x, y) in edge graph
(pixel value at the place x+ Δ x, y+ Δ y), p are adjustment factor and are positive integer mark.
10. eye fundus image joining method according to claim 9, which is characterized in that the k3=k4=3；
Or step S51 be establish it is one wide centered on scheme width k3Again, the high k of a height of Centered Graphs4Null images again are as fusion figure, institute
State k3、k4For the positive integer not less than 3, Centered Graphs are copied into fusion figure centre.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810980643.5A CN109255753B (en) | 2018-08-27 | 2018-08-27 | Fundus image splicing method |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810980643.5A CN109255753B (en) | 2018-08-27 | 2018-08-27 | Fundus image splicing method |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109255753A true CN109255753A (en) | 2019-01-22 |
| CN109255753B CN109255753B (en) | 2023-04-11 |
Family
ID=65049767
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810980643.5A Active CN109255753B (en) | 2018-08-27 | 2018-08-27 | Fundus image splicing method |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109255753B (en) |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114862760A (en) * | 2022-03-30 | 2022-08-05 | 中山大学中山眼科中心 | Method and device for detecting retinopathy of prematurity |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140198298A1 (en) * | 2013-01-14 | 2014-07-17 | Altek Corporation | Image stitching method and camera system |
| 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 |
| CN107292868A (en) * | 2017-05-31 | 2017-10-24 | 瑞达昇科技(大连)有限公司 | A kind of optic disk localization method and device |
| CN108022228A (en) * | 2016-10-31 | 2018-05-11 | 天津工业大学 | Based on the matched colored eye fundus image joining method of SIFT conversion and Otsu |
-
2018
- 2018-08-27 CN CN201810980643.5A patent/CN109255753B/en active Active
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| US20140198298A1 (en) * | 2013-01-14 | 2014-07-17 | Altek Corporation | Image stitching method and camera system |
| CN108022228A (en) * | 2016-10-31 | 2018-05-11 | 天津工业大学 | Based on the matched colored eye fundus image joining method of SIFT conversion and Otsu |
| 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 |
| CN107292868A (en) * | 2017-05-31 | 2017-10-24 | 瑞达昇科技(大连)有限公司 | A kind of optic disk localization method and device |
Non-Patent Citations (3)
| Title |
|---|
| YULIANG WANG .ETC: "Automatic fundus images based on SIFT feature", 《2010 3RD INTERNATIONAL CONGRESS ON IMAGE AND SIGNAL PROCESSING》 * |
| 焦晶萍等: "一种基于模板匹配法的眼底图像拼接方法", 《计算机技术与发展》 * |
| 王玉亮等: "基于SIFT特征的眼底图像自动拼接", 《中国图象图形学报》 * |
Cited By (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114862760A (en) * | 2022-03-30 | 2022-08-05 | 中山大学中山眼科中心 | Method and device for detecting retinopathy of prematurity |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109255753B (en) | 2023-04-11 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| US20230274577A1 (en) | Device and method with image matching | |
| TW569148B (en) | Method for locating facial features in an image | |
| CN104463159B (en) | A kind of image processing method and device for positioning iris | |
| JP2008113902A (en) | Apparatus and method for detecting eye opening degree | |
| WO2013087026A1 (en) | Locating method and locating device for iris | |
| KR102095723B1 (en) | Improvements in or relating to image processing | |
| JP2008234208A (en) | Facial region detection apparatus and program | |
| CN111710012B (en) | OCTA imaging method and device based on two-dimensional composite registration | |
| WO2020019286A1 (en) | Blepharoptosis detection method and system | |
| Ramli et al. | Feature‐based retinal image registration using D‐saddle feature | |
| CN106447640A (en) | Multi-focus image fusion method based on dictionary learning and rotating guided filtering and multi-focus image fusion device thereof | |
| CN110909631A (en) | Finger vein image ROI extraction and enhancement method | |
| CN109255753A (en) | A kind of eye fundus image joining method | |
| Reel et al. | Robust retinal image registration using expectation maximisation with mutual information | |
| CN111192280B (en) | Method for detecting optic disc edge based on local feature | |
| CN113197549A (en) | System for diagnosing diseases through face recognition technology | |
| He et al. | Biomedical image sequence analysis with application to automatic quantitative assessment of facial paralysis | |
| WO2017046377A1 (en) | Method and computer program product for processing an examination record comprising a plurality of images of at least parts of at least one retina of a patient | |
| CN114998353B (en) | System for automatically detecting vitreous opacity spot fluttering range | |
| Guan | Robust eye detection from facial image based on multi-cue facial information | |
| Kim et al. | Segmentation of optic nerve head using warping and RANSAC | |
| Parekar et al. | Automatic retinal image registration using fully connected vascular tree | |
| Sabaz et al. | ROI detection and vessel segmentation in retinal image | |
| CN111626974B (en) | Quality scoring method and device for coronary angiography image sequence | |
| Dasgupta et al. | An improved algorithm for eye corner detection |
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 |