CN106446805B - A kind of eyeground shine in optic cup dividing method and system - Google Patents
A kind of eyeground shine in optic cup dividing method and system Download PDFInfo
- Publication number
- CN106446805B CN106446805B CN201610810816.XA CN201610810816A CN106446805B CN 106446805 B CN106446805 B CN 106446805B CN 201610810816 A CN201610810816 A CN 201610810816A CN 106446805 B CN106446805 B CN 106446805B
- Authority
- CN
- China
- Prior art keywords
- point
- information point
- colour difference
- difference information
- weight
- 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
- 238000000034 method Methods 0.000 title claims abstract description 93
- 238000005452 bending Methods 0.000 claims abstract description 166
- 230000002792 vascular Effects 0.000 claims abstract description 149
- 210000003733 optic disk Anatomy 0.000 claims abstract description 73
- 210000004204 blood vessel Anatomy 0.000 claims description 43
- 230000004075 alteration Effects 0.000 claims description 16
- 238000004364 calculation method Methods 0.000 claims description 12
- 239000008280 blood Substances 0.000 claims description 10
- 210000004369 blood Anatomy 0.000 claims description 10
- 238000012545 processing Methods 0.000 claims description 10
- 238000004737 colorimetric analysis Methods 0.000 claims description 6
- 230000011218 segmentation Effects 0.000 abstract description 26
- 208000010412 Glaucoma Diseases 0.000 abstract description 18
- 238000003745 diagnosis Methods 0.000 abstract description 15
- 238000010586 diagram Methods 0.000 description 16
- 239000000284 extract Substances 0.000 description 6
- 230000008569 process Effects 0.000 description 6
- 238000005516 engineering process Methods 0.000 description 3
- 238000000605 extraction Methods 0.000 description 3
- 238000000926 separation method Methods 0.000 description 3
- 238000012795 verification Methods 0.000 description 3
- 230000008901 benefit Effects 0.000 description 2
- 230000008859 change Effects 0.000 description 2
- 230000000052 comparative effect Effects 0.000 description 2
- 230000000694 effects Effects 0.000 description 2
- 208000030533 eye disease Diseases 0.000 description 2
- 239000011159 matrix material Substances 0.000 description 2
- 210000001328 optic nerve Anatomy 0.000 description 2
- 206010061323 Optic neuropathy Diseases 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000005314 correlation function Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 238000001514 detection method Methods 0.000 description 1
- 230000004438 eyesight Effects 0.000 description 1
- 238000007429 general method Methods 0.000 description 1
- 238000007689 inspection Methods 0.000 description 1
- 208000020911 optic nerve disease Diseases 0.000 description 1
- 210000000056 organ Anatomy 0.000 description 1
- 230000004281 retinal morphology Effects 0.000 description 1
- 230000002194 synthesizing effect Effects 0.000 description 1
- 230000009466 transformation Effects 0.000 description 1
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06V—IMAGE OR VIDEO RECOGNITION OR UNDERSTANDING
- G06V40/00—Recognition of biometric, human-related or animal-related patterns in image or video data
- G06V40/10—Human or animal bodies, e.g. vehicle occupants or pedestrians; Body parts, e.g. hands
- G06V40/18—Eye characteristics, e.g. of the iris
-
- 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
- Engineering & Computer Science (AREA)
- Health & Medical Sciences (AREA)
- General Health & Medical Sciences (AREA)
- Ophthalmology & Optometry (AREA)
- Human Computer Interaction (AREA)
- Physics & Mathematics (AREA)
- General Physics & Mathematics (AREA)
- Multimedia (AREA)
- Theoretical Computer Science (AREA)
- Eye Examination Apparatus (AREA)
Abstract
The embodiment of the present invention provide a kind of eyeground shine in optic cup dividing method and system.The described method includes: obtaining the optic disk line in the photograph of eyeground, multiple colour difference information points in optic disk line region are obtained according to preset rules；The multiple colour difference information point is carried out curve fitting, obtains initial optic cup line, and region of search is arranged according to the initial optic cup line；Multiple vascular bending points in described search region are obtained, the multiple colour difference information point and the multiple vascular bending point are carried out curve fitting and obtain optic cup line.The system is for executing the above method.The dividing method and system of optic cup, fully utilize the colour difference information and vessel information in the photograph of eyeground in eyeground photograph provided in an embodiment of the present invention, improve optic cup segmentation volume accuracy, further improve the accuracy of diagnosis of glaucoma result.
Description
Technical field
The present embodiments relate to technical field of medical image processing, and in particular to a kind of eyeground shine in optic cup segmentation side
Method and system.
Background technique
Eyes are one of most important organs of human body, if eye disease is suffered from, then the quality of life of patient will will receive
Very big influence.Glaucoma is one kind of eye disease, it is a kind of based on the change of optic disk (also known as optic papilla) structure
Want the progressivity optic neuropathy of feature, this structure change is mainly shown as optic disk disk along being gradually reduced, this illustrate aixs cylinder and
Optic cell is being degenerated.Since the function that optic nerve loses can not be restored again, early detection and in time treatment are to save
Rescue a key of glaucoma patient eyesight.
According to a kind of important inspection for being oculist's diagnosis glaucoma, eyeground, which is shone, to be referred to eye-ground photography technology to eye on eyeground
The photo that bottom is taken a picture.Eye-ground photography technology is to be connected using special instrument such as digital camera with ophthalmoscope, by eyeground
Image is shown on computers, can be printed and be got off to be stored in case history, can be to do a comparison before and after re-treatment, it can be objective
Record eyeground Posterior pole retinal morphology variation, have preferable objectivity, repeatability and comparability.Eyeground is assessed to shine
Middle optic disk structure and optic cup structure are an important links of diagnosis of glaucoma, and optic disk is that optic nerve enters ocular region position
Entrance, in typical two-dimentional eyeground is shone, optic disk be in a bright elliptical region, and optic cup is central recessed of optic disk
It falls into.The variation of artificial observation optic disk structure and optic cup structure is a time-consuming process, and accuracy can be because of the water of observer
It puts down and changes, in the process, optic disk segmentation and optic cup segmentation are important a step.Wherein, optic disk segmentation has been achieved with
A large amount of high accuracies as a result, the segmentation result with ophthalmologist, oculist is very close.Comparatively, the algorithm of optic cup segmentation
It is also fewer.And the precision of optic cup segmentation, compared with optic disk segmentation, also there are also biggish distances.
In the prior art, mostly use C-means clustering method (C means clustering method) on removing the image after blood vessel greatly
Optic cup is extracted by fuzzy C-mean algorithm wavelet transformation, or center is calculated from super-pixel statistical data using super-pixel classification method, and
Utilize histogram divion optic cup.Both methods is all finally to be fitted using ellipse to cup boundary, and all do not fill
Divide and utilize vessel information, thus be easy to miss the local feature of optic cup, to influence the accuracy of diagnosis of glaucoma result.
Therefore, how to propose a kind of method, reasonably optic cup can be split, improve the standard of diagnosis of glaucoma result
Exactness becomes urgent problem to be solved.
Summary of the invention
For the defects in the prior art, the embodiment of the present invention provide a kind of eyeground shine in optic cup dividing method and be
System.
On the one hand, the embodiment of the present invention propose a kind of eyeground shine in optic cup dividing method, comprising:
The optic disk line in the photograph of eyeground is obtained, obtains multiple colour difference informations in optic disk line region according to preset rules
Point；
The multiple colour difference information point is carried out curve fitting, obtains initial optic cup line, and according to the initial optic cup line
Region of search is set；
Multiple vascular bending points in described search region are obtained, by the multiple colour difference information point and the multiple blood
Pipe bending point, which carries out curve fitting, obtains optic cup line.
On the other hand, the embodiment of the present invention provide a kind of eyeground shine in optic cup segmenting system, comprising:
Module is obtained, for obtaining the optic disk line in the photograph of eyeground, is obtained in optic disk line region according to preset rules
Multiple colour difference information points；
Processing module obtains initial optic cup line, and according to institute for the multiple colour difference information point to carry out curve fitting
State initial optic cup line setting region of search；
Fitting module, for obtaining multiple vascular bending points in described search region, by the multiple colour difference information point
And the multiple vascular bending point carries out curve fitting and obtains optic cup line.
The dividing method and system of optic cup, fully utilize the color in the photograph of eyeground in eyeground photograph provided in an embodiment of the present invention
Poor information and vessel information obtain colour difference information point and vascular bending point, fit more accurate optic cup line, and doctor is true
The process modelling for determining optic cup line in the photograph of eyeground, is not only utilized the position of colour difference information point and vascular bending point, also utilizes
Each colour difference information point and the corresponding reliability of vascular bending point, and reasonably determined each colour difference information point with
And the weight of vascular bending point, it carries out curve fitting in conjunction with weight, so that even if having a small number of colour difference information points or vascular bending point
Inaccuracy still is able to obtain accurate optic cup line.The accuracy for improving optic cup line fitting result reduces artificial progress
The amount of labour of optic cup segmentation, further improves the accuracy of diagnosis of glaucoma result.
Detailed description of the invention
In order to more clearly explain the embodiment of the invention or the technical proposal in the existing technology, to embodiment or will show below
There is attached drawing needed in technical description to be briefly described, it should be apparent that, the accompanying drawings in the following description is this hair
Bright some embodiments for those of ordinary skill in the art without creative efforts, can be with root
Other attached drawings are obtained according to these attached drawings.
Fig. 1 is the flow diagram of the dividing method of optic cup in the photograph of eyeground in the embodiment of the present invention；
Fig. 2 is the idiographic flow schematic diagram of the dividing method of optic cup in the photograph of eyeground in the embodiment of the present invention；
Fig. 3 is the flow diagram of the dividing method of optic cup in another eyeground photograph in the embodiment of the present invention；
Fig. 4 is the dividing method comparative diagram of optic cup in the photograph of eyeground in the embodiment of the present invention；
Fig. 5 is the structural schematic diagram of the segmenting system of optic cup in the photograph of eyeground in the embodiment of the present invention；
Fig. 6 is the structural schematic diagram of the segmenting system of optic cup in another eyeground photograph in the embodiment of the present invention；
Fig. 7 is the structural schematic diagram of the segmenting system of optic cup in another eyeground photograph in the embodiment of the present invention；
Fig. 8 is the structural schematic diagram of the segmenting system of optic cup in another eyeground photograph in the embodiment of the present invention.
Specific embodiment
In order to make the object, technical scheme and advantages of the embodiment of the invention clearer, below in conjunction with the embodiment of the present invention
In attached drawing, technical scheme in the embodiment of the invention is clearly and completely described, it is clear that described embodiment is
A part of the embodiment of the present invention, instead of all the embodiments.Based on the embodiments of the present invention, those of ordinary skill in the art
Every other embodiment obtained without creative efforts, shall fall within the protection scope of the present invention.
Fig. 1 is the flow diagram of the dividing method of optic cup in the photograph of eyeground in the embodiment of the present invention, as shown in Figure 1, this reality
Apply example disclose a kind of eyeground shine in optic cup dividing method, comprising:
Optic disk line in S10, acquisition eyeground photograph, obtains the letter of multiple color difference in optic disk line region according to preset rules
Breath point；
Specifically, first by eyeground according in the segmenting system of optic cup in incoming eyeground photograph, shining medium vessels due to eyeground can be right
Colour difference information point is obtained to interfere, the embodiment of the present invention first by eyeground shine in blood vessel erase, then to eyeground according to image
Red value carries out optic disk segmentation using CV (Chan-Vese active contour model, CV active contour model) algorithm,
Obtain the optic disk line in the photograph of eyeground.It, can be according to preset rules since color difference is larger inside and outside near optic cup line
Multiple colour difference information points are got in optic disk line region, colour difference information point refers to the biggish separation of color difference in optic disk line.
Wherein red value refers to the R value in photochrome in rgb value.It is understood that can also be to eyeground when carrying out optic disk segmentation
Blue valve or green value according in carry out optic disk segmentation, and the effect that red value carries out optic disk segmentation is relatively good, the embodiment of the present invention
It is preferred that optic disk segmentation is carried out to the red value in the photograph of eyeground, further, it is also possible to obtain the optic disk in the photograph of eyeground using other methods
Line is not limited to the method enumerated in the embodiment of the present invention.
S11, the multiple colour difference information point is carried out curve fitting, obtains initial optic cup line, and according to the initial view
Region of search is arranged in cup line；
Specifically, after getting multiple colour difference information points in optic disk line, using computer software by the multiple color difference
Information point is fitted to the curve of a closure, as initial optic cup line.The closed curve fitting under rectangular coordinate system is asked first
The normal profile fitting problems being converted under polar coordinate system are inscribed, the colour difference information point that particularly will acquire carries out coordinate and turns
It changes, i.e., rectangular co-ordinate is converted into polar coordinates, carries out curve fitting to obtain a song by these colour difference information points under polar coordinate system
Line, then it is converted into rectangular co-ordinate by polar coordinates are carried out, the closed curve under rectangular coordinate system, that is, initial optic cup line are obtained,
And region of search is set near initial optic cup line, to search for vascular bending point.Wherein specifically multiple colour difference information points are intended
Synthesizing a closed curve can be fitted to obtain using regularization curve matching by the discrete coordinates data of multiple colour difference information points
Function data can specifically use regularization least square approximating method, so that the curve that fitting obtains is more smooth reasonable.
It should be noted that multiple colour difference information points can also carry out curve fitting using other methods, it is not limited to implementation of the present invention
The approximating method that example is enumerated, such as: utilize computer software Curve Fitting Toolbox etc..
S12, multiple vascular bending points in described search region are obtained, by the multiple colour difference information point and described more
A vascular bending point, which carries out curve fitting, obtains optic cup line.
Specifically, due to that can occur more significantly to be bent when blood vessel passes through optic cup line, it is arranged near initial optic cup line
Region of search in it is available arrive multiple vascular bending points, then multiple vascular bending points and multiple colour difference informations point are fitted to
The curve of one closure, as optic cup line, wherein vascular bending point refers to the biggish separation of vascular bending.Specific curve is quasi-
Conjunction method is identical as the above-mentioned method that multiple colour difference information points are fitted to initial optic cup line, and details are not described herein again.In the field of search
The method that vascular bending point is obtained in domain, which can be, first extracts all blood vessels in eyeground is shone, then extract on every blood vessel bending compared with
Apparent separation is vascular bending point, finally obtains the vascular bending point in region of search；It can also directly extract
Blood vessel in region of search, then the vascular bending point on the blood vessel in region of search is extracted, when obtaining vascular bending point certainly
The blood vessel erased should be restored first, or blood vessel of not erasing at another again identical eyeground shine in obtain blood vessel
Bending point, then by two open one's eyes bottom shine in data summarization to it is same open one's eyes bottom shine in.The specific method for obtaining vascular bending point can
To be, the vessel centerline of the blood vessel searched is first extracted, takes a point every 5 pixels on heart line in the blood vessel, is calculated
The vascular bending angle of each position will be put as blood accordingly when bending angle is less than a predetermined angle threshold value
Pipe bending point can also be other methods certainly to determine vascular bending point, and the embodiment of the present invention is not especially limited.It needs
Bright, the calculating of vascular bending angle can use relevant formula or be counted using the correlation function in computer software
It calculates, specific calculation method is not specifically limited herein, in addition, predetermined angle threshold value can also be carried out according to actual use situation
Setting, the embodiment of the present invention are not especially limited.
The dividing method of optic cup, the optic cup line fitted have both merged color difference in eyeground photograph provided in an embodiment of the present invention
Information point, and vascular bending point is combined, than only applying the ellipse fitting method of colour difference information and only using the slotting of vascular bending point
Value method embodies apparent advantage, so that the optic cup line obtained is more accurate reasonable, further improves diagnosis of glaucoma result
Accuracy.
On the basis of the above embodiments, described to obtain the letter of multiple color difference in optic disk line region according to preset rules
Breath point, comprising:
The center for obtaining the corresponding region of the optic disk line is made one every predetermined angle as endpoint using the center and is penetrated
Line obtains the ray of respective numbers；
A pixel is taken every presetted pixel on each ray, is calculated along each pixel described in the directions of rays
Value of chromatism inside and outside corresponding takes the maximum aberration on the ray to be worth corresponding pixel as the institute on the ray
Colour difference information point is stated, wherein the pixel is in optic disk line region.
Specifically, the center of optic disk line corresponding region is obtained, and makees one by each predetermined angle of endpoint of the center
Ray thus obtains a plurality of ray, and wherein the center of optic disk line corresponding region refers to the geometric center of optic disk line.Each
On ray, a pixel is taken every presetted pixel, and pixel is in optic disk line, that is to say, that be in optic disk in ray
A pixel is taken every presetted pixel on line segment in line region.All pixels point is calculated along the direction of ray where it
The value of chromatism of outer two sides, i.e. calculating all pixels point where it ray from the color difference inside and outside ray endpoint outward direction
Maximum aberration on every ray is worth corresponding pixel as colour difference information point by value.Wherein predetermined angle and presetted pixel can
To preset as needed, the embodiment of the present invention is not specifically limited.
Such as: after the geometric center for getting optic disk line, using the center as endpoint, every 10 degree of works, one ray, this
Sample just obtains 36 rays.A pixel is taken every 5 pixels in the part that every ray is in optic disk line, is calculated every
For a pixel along the interior outside value of chromatism of directions of rays where it, specific calculation method, which may is that, calculates each pixel institute
The average brightness of adjacent 5 pixels on ray from pixel to ray endpoint direction and from pixel along ray outward
To adjacent 5 pixels average brightness difference, the as corresponding value of chromatism of pixel, when the last one pixel is corresponding
When along the outside of directions of rays where it without 5 pixels, such as only 3 pixels can choose and abandon this pixel, can also
Only to calculate the average brightness of this 3 pixels.Taking the corresponding pixel of maximum aberration value on every ray is colour difference information
Point thus obtains 36 colour difference information points.It should be noted that predetermined angle is 10 degree in the embodiment of the present invention, just
Optic disk line can be averagely divided into 36 parts, predetermined angle is also possible to other degrees, such as 5 degree, 36 degree, can be according to reality
It needs to be arranged, the predetermined angle that also will appear setting certainly cannot averagely divide optic disk line, and such as 11 degree, such situation is fair
Perhaps, the ray obtained in this way is 33, is certainly preferably to be configured to be divided exactly by 360 by predetermined angle, is achieved in that
Colour difference information point it is more uniform and accurate.In addition, the calculation method of value of chromatism can also be calculated using other methods, it is not limited to
The calculation method enumerated in the embodiment of the present invention.
Optic cup dividing method in eyeground photograph provided in an embodiment of the present invention, by drawing a plurality of ray in optic disk line,
And the value of chromatism inside and outside directions of rays of the pixel on ray is calculated, take the corresponding pixel of maximum aberration value as color
Poor information point, to fit initial optic cup line.The optic cup line larger feature of interior outside color difference nearby is utilized, so that obtaining
The initial optic cup line got is more accurate reasonable, and progress can fit more accurate optic cup line, examine to improve glaucoma
The accuracy of disconnected result.
On the basis of the above embodiments, described to include: according to the initial optic cup line setting region of search
Institute where calculating the corresponding maximum aberration value of the colour difference information point and the colour difference information point on ray
There is the ratio of the color difference average value of the pixel, using the ratio as corresponding first reliability of the colour difference information point；
The inside curve and outside curve of the initial optic cup line, the inside are respectively set according to first reliability
Region between curve and the outside curve is described search region.
Specifically, after getting colour difference information point, the color difference of all pixels point where calculating colour difference information point on ray
The average value of value, using the ratio of the corresponding maximum aberration value of colour difference information point and the average value as the colour difference information point pair
The first reliability answered, i.e. corresponding first reliability of a colour difference information point, thus obtains multiple first reliabilitys.
Using the center of optic disk line as origin, a plurality of radius of inside curve is set along above-mentioned each ray according to the first reliability, i.e.,
The radius of the corresponding inside curve of one ray, wherein the radius of inside curve and the first reliability are linear, and
The maximum radius of inside curve cannot be less than the half of initial optic cup line maximum radius, the radius of all inside curves it is non-described
The curve that the endpoint at center fits is inside curve.Outside curve is obtained using identical method, and outside curve
Maximum radius no more than disk along the 4/5 of width, disk along the part referred between optic disk line and initial optic cup line, inside curve and
Region between the curve of outside is region of search.The wherein linear relationship of the radius of inside curve and the first reliability such as formula
(1) shown in, shown in the linear relationship formula such as formula (2) of the radius of outside curve and the first reliability, it is to be understood that interior
The radius of lateral curvature line and the radius of outside curve can also be obtained by other methods, be not limited to arrange in the embodiment of the present invention
The method of act.It should be noted that inside curve and outside curve can also be obtained by other methods, and such as: initially to regard
Cup line is boundary, according to the first reliability respectively to search width on the inside of setting on the inside of initial optic cup line on above-mentioned ray, to
Setting outside search width, the side endpoint of inside search width are fitted to obtain inside curve on the outside of initial optic cup line, and outside is searched
The side endpoint of Suo Kuandu is fitted to obtain outside curve.
RInside=RCup(1/2+1/2 (the first reliability-min)/(max-min)) (1)
ROutside=RCup+ 1.4 (4/5-3/5 (the first reliability-min)/(max-min)) RDisk edge (2)
In formula: min refers to the first reliability of minimum in the first reliability of all colour difference information points；Max refers to all
The first reliability of maximum in first reliability of colour difference information point；RDisk edgeRefer to the radius and the half of initial optic cup line of optic disk line
The difference of diameter.It should be noted that the linear relationship formula that the embodiment of the present invention is enumerated is the preferred formula by experimental verification, tool
Body can also be not especially limited when using using other linear formulas, the embodiment of the present invention.
Below by the method that specific example introduces region of search acquisition, such as: method through the foregoing embodiment exists
36 rays and 36 colour difference information points are acquired in optic disk line, it is assumed that it is corresponding that one of colour difference information point A is calculated
The color difference average value of all pixels point on ray is 6, and the corresponding maximum aberration value of colour difference information point A is 8, then colour difference information
The first reliability of point A is 4/3, it is assumed that the radius of the optic disk line of acquisition is RDisk, the radius of initial optic cup line is RCup, it is calculated
The first reliability of minimum be 7/6, maximum first reliability is 5/3, then A pairs of colour difference information point can be obtained according to formula (1)
The radius R for the initial optic disk line inside curve answeredInside, the radius R of outside curve is obtained according to formula (1)Outside.With optic disk line center
It is origin respectively with RInsideAnd ROutsideMake the corresponding interior lateral curvature of color difference information point A on the ray where colour difference information point A for radius
The radius of line and the radius of outside curve, acquire two terminal As accordingly1And A2.Same method gets other color difference letter
The endpoint of the radius of the endpoint and outside curve of the radius of the corresponding inside curve of breath point, respectively by the half of this 36 inside curves
The endpoint of diameter and the endpoint of outside sweep are fitted, and obtain inside curve and outside curve, and region of search is interior lateral curvature
Part between line and outside curve.
The dividing method of optic cup, passes through and calculates the first of each colour difference information point in eyeground photograph provided in an embodiment of the present invention
Reliability, and inside curve and outside curve are obtained according to the first reliability, reasonable region of search is obtained, is enabled to
Vascular bending point is more accurately searched for and got, optic cup line is obtained further combined with colour difference information point, improves optic cup point
The accuracy cut, and then improve the accuracy of diagnosis of glaucoma result.
On the basis of the above embodiments, by the multiple colour difference information point and the multiple vascular bending point march
Line fitting obtains optic cup line
Calculate corresponding multiple second reliabilitys of the multiple vascular bending point；
According to corresponding multiple first weights of the multiple colour difference information point of first reliability calculating, according to described
Corresponding multiple second weights of the multiple vascular bending point of two reliability calculatings；
According to the multiple first weight and the multiple second weight by the multiple colour difference information point and described more
A vascular bending point, which carries out curve fitting, obtains optic cup line.
Specifically, the second reliability of the vascular bending point in region of search is calculated first, and a vascular bending point is corresponding
One the second reliability can according to second according to corresponding multiple first weights of the multiple colour difference information points of the first reliability calculating
Calculate corresponding multiple second weights of multiple vascular bending points by spending, i.e. corresponding first weight of a colour difference information point, one
Corresponding second weight of a vascular bending point.The first weight of specific calculating and the second weight can be with the following method:
Firstly, total weight distribution: regarding all colour difference information points as first information point, all vascular bending points are seen
Make the second information point, total weight 1 is distributed into first information point and the second information point.Total weight 1 is specifically distributed into the first letter
The method of breath point and the second information point can refer to following steps:
R1, the average value for calculating corresponding all first reliabilitys of above-mentioned all colour difference information points are flat as color difference ratio
Mean value；
If R2, color difference average of relatives value are more than or equal to 3, enabling the weight of first information point is 2/3, the power of the second information point
Weight is 1/3；
If R3, color difference average of relatives value are less than or equal to 2, enabling the weight of first information point is 1/3, the power of the second information point
Weight is 2/3；
If it is 1/2 that R4, color difference average of relatives value, which between 2 and 3, enable the weight of first information point, the second information point
Weight be 1/2.
Secondly, the first weight and the second weight distribution: the corresponding weight distribution of first information point is believed to each color difference
Point, i.e. each corresponding first weight of colour difference information point are ceased, wherein the first weight and the first reliability are linear, specifically
Linear relationship such as formula (3) shown in, by the corresponding weight distribution of the second information point give each vascular bending point, the second weight
It is linear with the second reliability, shown in specific linear relationship such as formula (4).It is understood that the embodiment of the present invention
In the numerical value enumerated such as: 2,3,1/3 is the preferred value by experimental verification, can be according to service condition when practical application
It is configured, the embodiment of the present invention is not specifically limited.
First weight=first information point weight * (the sum of the first reliability/first reliability) (3)
Second weight=second information point weight * (the sum of the second reliability/second reliability) (4)
In formula: the weight of first information point refers in total weight distribution, distributes to the weight of all colour difference information points, the
The weight of two information points refers in total weight distribution, distributes to the weight of all vascular bending points, such as: if calculating gained color
Poor average of relatives value is more than or equal to 3, then the weight of first information point is 2/3, and the weight of the second information point is 1/3；First is reliable
Degree refers to that corresponding first reliability of the corresponding colour difference information point of the first weight, the sum of first reliability refer to all colour difference informations
The sum of corresponding all first reliabilitys of point, the sum of corresponding second reliability and the second reliability are similar meanings, only
It is that colour difference information point is changed to vascular bending point, details are not described herein again.It should be noted that the embodiment of the present invention enumerate it is linear
Relation formula (3) and formula (4) are the preferred formulas by experimental verification, can also be using other linear public when specifically used
Formula, the embodiment of the present invention are not especially limited.
The method to carry out curve fitting with specific reference to the first weight and the second weight can be, by each first weight and
Each second weight is rounded multiplied by default value, such as 100, and by acquired results, as corresponding colour difference information point and phase
The number that the vascular bending point answered occurs, what the number and each vascular bending point occurred in conjunction with each colour difference information point occurred
All colour difference information points and all vascular bending points are carried out curve fitting, obtain optic cup line by number.Such as: assuming that having 3
A colour difference information point a, b, c, corresponding transverse and longitudinal coordinate matrix and ordinate matrix are x=[x1, x2, x3] respectively, y=[y1,
Y2, y3], it is assumed that corresponding first weight of a, b, c is 0.01,0.02,0.03 respectively, by the first weight respectively multiplied by after 100
To 1,2,3, then it represents that a, b, c occur respectively 1 time, twice, three times.In conjunction with the coordinate value of colour difference information point when carrying out curve fitting
With the number for calculating resulting each color difference information point and occurring, when curve matching the coordinate representation of colour difference information point be x=[x1, x2,
X2, x3, x3, x3], y=[y1, y2, y2, y3, y3, y3], the same method of vascular bending point.It should be noted that according to
One weight and the second weight, which carry out curve fitting to colour difference information point and vascular bending point, to be not limited to arrange in the embodiment of the present invention
The method of act, can also use other methods, for example, using in computer software relevant weight fitting function or weight it is quasi-
Tool box is closed to carry out curve fitting.
The dividing method of optic cup, calculates each colour difference information point and blood vessel in eyeground photograph provided in an embodiment of the present invention
The corresponding weighted value of bending point comes to carry out curve to the colour difference information point and vascular bending point of acquisition in conjunction with corresponding weighted value
Fitting still can obtain accurate view in the case where there is the colour difference information point or vascular bending point of a small number of inaccuracy
Cup line, improves the anti-noise ability of curve matching, so that the optic cup line that fitting obtains is more accurate, further increases green light
The accuracy of eye diagnostic result.
On the basis of the above embodiments, described to calculate the corresponding multiple second reliability packets of the multiple vascular bending point
It includes: obtaining the corresponding vascular bending curvature of each vascular bending point and blood vessel diameter, the vascular bending curvature and corresponding
The product of the blood vessel diameter is the second reliability of the corresponding vascular bending point.
Specifically, corresponding vascular bending point in region of search is obtained first, and obtains the corresponding coordinate of vascular bending point
Value, calculates the corresponding vascular bending curvature of the vascular bending point, and the specific method for calculating vascular bending curvature can use blood
The correlation formula that pipe bending curvature calculates is calculated, and also can use the phase of the vascular bending curvature estimation in computer software
Function is closed to calculate, other methods can also be used certainly, the specific calculation method embodiment of the present invention is not especially limited.Again
The blood vessel diameter at vascular bending point is obtained, the specific method for obtaining blood vessel diameter can be, using vascular bending point as the center of circle, half
Picture circle is gradually expanded in diameter, and stopping when intersecting with vessel borders, corresponding circular diameter is that blood vessel is straight at vascular bending point at this time
Diameter.Using obtained vascular bending curvature and the product of corresponding blood vessel diameter as the second reliable of corresponding vascular bending point
Degree.It can be understood that the acquisition methods of vascular bending curvature and blood vessel diameter are not limited to the side enumerated in the embodiment of the present invention
Method.
The dividing method of optic cup, combines colour difference information point and vascular bending in eyeground photograph provided in an embodiment of the present invention
Point, and calculate colour difference information point and the corresponding weighted value of vascular bending point, in conjunction with weighted value by the colour difference information point of acquisition
It carries out curve fitting with vascular bending point, obtains optic cup line, avoid a small number of colour difference information points or vascular bending point inaccuracy shadow
The accuracy that optic cup line obtains is rung, the accuracy of optic cup segmentation is improved, so that the accuracy of diagnosis of glaucoma result is improved,
It reduces artificial observation and carries out the amount of labour of optic cup segmentation, while it is accurate to avoid the level difference optic cup due to observer from dividing
Degree is different, influences the accuracy of diagnosis of glaucoma result.
The specific dividing method of optic cup in the photograph of eyeground is described below, so that technical solution of the present invention is better understood,
Fig. 2 is the idiographic flow schematic diagram of the dividing method of optic cup in the photograph of eyeground in the embodiment of the present invention, and Fig. 3 is in the embodiment of the present invention
The flow diagram of the dividing method of optic cup regards in the photograph of eyeground in the embodiment of the present invention as shown in Figure 2 and Figure 3 in another eyeground photograph
Specific step is as follows for the dividing method of cup:
Blood vessel in S20, eyeground photograph of erasing.Since when determining colour difference information point, the blood vessel in the photograph of eyeground will cause dry
Disturb, therefore, first by eyeground shine in blood vessel erase, obtain without blood vessel eyeground shine.
S21, optic disk line is extracted.The method that the embodiment of the present invention extracts optic disk line is that the red value in shining eyeground utilizes CV
Algorithm determines the optic disk line in eyeground, and other methods can also be used when specifically used.
S22, colour difference information point, the initial optic cup line obtained are obtained.The specific method for obtaining colour difference information point can be as above-mentioned
The method introduced in embodiment is drawn a plurality of ray by origin of the center of optic disk line, is taken respectively on each ray multiple
Pixel, and the corresponding directions of rays where it of the pixel on each ray is calculated by optic disk line center outward direction
Value of chromatism takes on each ray the corresponding pixel of maximum aberration value as the corresponding colour difference information point of the ray, and counts
Each corresponding first reliability of colour difference information point is calculated, calculates the method for the first reliability with the method one in above-described embodiment
It causes, details are not described herein again.
S23, region of search is determined.The inside curve and outer lateral curvature of initial optic disk line are respectively set according to the first reliability
Line, the part between inside curve and outside curve are region of search, the setting method of inside curve and outside curve and upper
The method stated in embodiment is consistent, and details are not described herein again.
All vascular bending points in S24, extraction eyeground photograph.The blood vessel in the photograph of eyeground is obtained, and extracts the blood on blood vessel
Pipe bending point.
Vascular bending point in S25, extraction region of search.It is being extracted in extraction step S24 and determining in step S23
Region of search in vascular bending point, and calculate each corresponding second reliability of vascular bending point, specifically extract blood vessel
Bending point and the second reliability of calculating method is consistent with above-mentioned fact Example, and details are not described herein again.
S26, curve matching obtain optic cup line.According to the first reliability and the second reliability calculate separately colour difference information point and
The first weight and the second weight of vascular bending point, using the first weight and the second weight multiplied by default value as corresponding color difference
The number that information point and vascular bending point occur, according to calculating resulting number for colour difference information point and vascular bending point march
Line fitting, first carries out coordinate conversion, and rectangular coordinate system is converted to polar coordinate system, and fitting obtains the optic cup line under polar coordinate system,
Polar coordinate system is converted into rectangular coordinate system again, obtains the optic cup line under rectangular coordinate system, specific fit procedure and above-mentioned implementation
Example is consistent, and details are not described herein again.
Fig. 4 is the dividing method comparative diagram of optic cup in the photograph of eyeground in the embodiment of the present invention, as shown in Figure 4, it can be seen that this
The dividing method of optic cup has a significant effect with the comparison of general method in the eyeground photograph that inventive embodiments provide, and the present invention is real
It applies the resulting optic cup segmentation result of example and ophthalmologist, oculist carries out the result of manpower optic cup segmentation with high consistency.
The dividing method of optic cup, fully utilizes the colour difference information in the photograph of eyeground in eyeground photograph provided in an embodiment of the present invention
Colour difference information point and vascular bending point are obtained with vessel information, more accurate optic cup line is fitted, doctor is determined into eyeground
According to the process modelling of middle optic cup line, the position of colour difference information point and vascular bending point is not only utilized, also uses each
The reliability degree of colour difference information point and vascular bending point, and reasonably it has been determined that each colour difference information point and blood vessel are curved
The weight of song point, carries out curve fitting in conjunction with weight, so that even if having a small number of colour difference information points or vascular bending point inaccuracy, still
Accurate optic cup line can be obtained.The accuracy for improving optic cup line fitting result reduces artificial progress optic cup segmentation
The amount of labour, further improve the accuracy of diagnosis of glaucoma result.
Fig. 5 is the structural schematic diagram of the segmenting system of optic cup in the photograph of eyeground in the embodiment of the present invention, as shown in figure 5, this hair
Bright embodiment offer is that the segmenting system of optic cup during eyeground is shone includes: to obtain module 50, processing module 51 and fitting module 52,
Wherein:
Module 50 is obtained for obtaining the optic disk line in the photograph of eyeground, is obtained in optic disk line region according to preset rules
Multiple colour difference information points；Processing module 51 is used to carry out curve fitting the multiple colour difference information point, obtains initial optic cup line,
And region of search is arranged according to the initial optic cup line；Multiple blood vessels that fitting module 52 is used to obtain in described search region are curved
The multiple colour difference information point and the multiple vascular bending point are carried out curve fitting and obtain optic cup line by Qu Dian.
Specifically, obtain module 50 by eyeground shine in blood vessel erase, and obtain eyeground shine in optic disk line, further according to pre-
If multiple colour difference information points in Rule optic disk line region, and colour difference information point is sent to processing module 51；Handle mould
Block 51 carries out curve fitting colour difference information point, obtains initial optic cup line, and determine region of search；The acquisition of fitting module 52 is being located
The vascular bending point in the region of search that module 51 determines is managed, and acquired vascular bending point and acquisition module 50 are obtained
Colour difference information point carry out curve fitting to obtain optic cup line.The specific method for obtaining optic disk line and curve matching and above-mentioned of obtaining
Embodiment is consistent, and details are not described herein again.
The segmenting system of optic cup in eyeground photograph provided in an embodiment of the present invention, has merged the colour difference information and blood in the photograph of eyeground
Pipe bending information, by doctor determine eyeground shine in optic cup line process modelling, improve eyeground shine in optic cup segmentation it is accurate
Property, the accuracy of diagnosis of glaucoma result is further improved, while reducing the artificial amount of labour for carrying out optic cup segmentation, avoided
Artificial segmentation as it is horizontal it is different caused by optic cup segmentation result it is different.
Fig. 6 is the structural schematic diagram of the segmenting system of optic cup in another eyeground photograph in the embodiment of the present invention, as shown in fig. 6,
On the basis of the above embodiments, the acquisition module 50 further includes ray acquisition submodule 501 and Colorimetry submodule
502, in which:
Ray acquisition submodule 501 is used to obtain the center in the corresponding region of the optic disk line, using the center as endpoint
Make a ray every predetermined angle, obtains the ray of respective numbers；Colorimetry submodule 502 is used in each ray
Take a pixel every presetted pixel, calculate along each pixel described in the directions of rays it is corresponding inside and outside color difference
Value takes the maximum aberration on the ray to be worth corresponding pixel as the colour difference information point on the ray, wherein institute
Pixel is stated in optic disk line region.
Specifically, ray acquisition submodule 501 gets the center of optic disk line, and draws from the center every predetermined angle
A ray out obtains the ray of respective numbers, and the specific method for obtaining ray is consistent with above-described embodiment, no longer superfluous herein
It states.Colorimetry submodule 502 takes a pixel every presetted pixel on the ray that ray acquisition submodule 501 obtains,
And the pixel calculates endpoint outward direction of each pixel along directions of rays where it by ray in optic disk line
Inside and outside value of chromatism, take on each ray the corresponding pixel of maximum aberration value as colour difference information point.Wherein preset
The calculating of angle, presetted pixel and value of chromatism is all consistent with above-described embodiment, and details are not described herein again.
Optic cup segmenting system in eyeground photograph provided in an embodiment of the present invention is made by reasonably determining colour difference information point
The initial optic cup line that must be got is more accurate reasonable, is further able to fit more accurate optic cup line, to improve blueness
The accuracy of light eye diagnostic result.
Fig. 7 is the structural schematic diagram of the segmenting system of optic cup in another eyeground photograph in the embodiment of the present invention, as shown in fig. 7,
On the basis of the above embodiments, the processing module 51 further includes that the first reliability calculating submodule 511 and region of search are true
Stator modules 512, in which:
First reliability calculating submodule 511 is for calculating the corresponding maximum aberration value of the colour difference information point and institute
The ratio of the color difference average value of all pixels where stating colour difference information point on ray, using the ratio as the color
Corresponding first reliability of poor information point；Region of search determines submodule 512, for being respectively set according to first reliability
The inside curve and outside curve of the initial optic cup line, the region between the inside curve and the outside curve is described
Region of search.
Specifically, after Colorimetry submodule 502 gets colour difference information point, the first reliability calculating submodule 511 meter
The average value of the value of chromatism of all pixels point where calculating colour difference information point on ray, by the corresponding maximum aberration of colour difference information point
Value is with the ratio of the average value as corresponding first reliability of the colour difference information point.Region of search determines submodule 512
The inside curve and outside curve for determining initial optic cup line respectively further according to the first reliability, between inside curve and outside curve
Region be described search region.It is specific to determine that inside curve is consistent with above-described embodiment with the method for outside curve, herein not
It repeats again.
The segmenting system of optic cup, the reasonable field of search for determining vascular bending point in eyeground photograph provided in an embodiment of the present invention
Domain obtains optic cup line further combined with vascular bending point and colour difference information point, improves the accuracy of optic cup segmentation, and then improve
The accuracy of diagnosis of glaucoma result.
Fig. 8 is the structural schematic diagram of the segmenting system of optic cup in another eyeground photograph in the embodiment of the present invention, as shown in fig. 7,
On the basis of the above embodiments, the fitting module 52 further includes the second reliability calculating submodule 521, weight calculation submodule
Block 522 and curve matching submodule 523, in which:
Second reliability calculating submodule 521 is for calculating the multiple vascular bending point corresponding multiple second reliably
Degree；Weight calculation submodule 522 is used for according to the multiple colour difference information point of first reliability calculating corresponding multiple the
One weight, according to corresponding multiple second weights of the multiple vascular bending point of second reliability calculating；Curve matching
Module 523 is used for the multiple colour difference information point and described according to the multiple first weight and the multiple second weight
Multiple vascular bending points, which carry out curve fitting, obtains optic cup line.
Specifically, the second reliability calculating submodule 521 gets the point of the vascular bending in region of search and calculates search
Corresponding second reliability of vascular bending point in region, corresponding second reliability of a vascular bending point.Weight calculation
Submodule 522 is according to corresponding multiple first weights of the multiple colour difference information points of the first reliability calculating, according to the second reliability meter
Calculate corresponding multiple second weights of multiple vascular bending points, i.e. corresponding first weight of a colour difference information point, a blood vessel
Bending point corresponds to second weight.Wherein the calculation method of the first weight and the second weight is consistent with above-described embodiment herein
It repeats no more.Curve matching submodule 523 is according to the first weight and the second weight by all colour difference information points and all blood
Pipe bending point carries out curve fitting, and obtains optic cup line, specific approximating method is consistent with above-described embodiment, and details are not described herein again.
The segmenting system of optic cup, calculates each colour difference information point and blood vessel in eyeground photograph provided in an embodiment of the present invention
The corresponding weighted value of bending point, so that in the case where there is the colour difference information point or vascular bending point of a small number of inaccuracy, still
Accurate optic cup line can be obtained, the anti-noise ability of curve matching is improved, so that the optic cup line that fitting obtains is more
Accurately, the accuracy of diagnosis of glaucoma result is further increased.
On the basis of the above embodiments, the second reliability calculating submodule 521 is specifically used for: it is curved to obtain each blood vessel
Song puts corresponding vascular bending curvature and blood vessel diameter, and the product of the vascular bending curvature and the corresponding blood vessel diameter is
Second reliability of the corresponding vascular bending point.
Specifically, the second reliability calculating submodule 521 obtains corresponding vascular bending point in region of search, and obtains blood
The corresponding coordinate value of pipe bending point calculates the blood vessel at the corresponding vascular bending curvature of vascular bending point and vascular bending point
The product of diameter, vascular bending curvature and the corresponding blood vessel diameter is the second reliable of the corresponding vascular bending point
Degree.The specific method for calculating the blood vessel diameter at the corresponding vascular bending curvature of vascular bending point and vascular bending point is the same as above-mentioned
Embodiment is consistent, and details are not described herein again.
The segmenting system of optic cup, fully utilizes the colour difference information in the photograph of eyeground in eyeground photograph provided in an embodiment of the present invention
Colour difference information point and vascular bending point are obtained with vessel information, more accurate optic cup line is fitted, doctor is determined into eyeground
According to the process modelling of middle optic cup line, the position of colour difference information point and vascular bending point is not only utilized, also uses each
Colour difference information point and the corresponding reliability degree of vascular bending point, and each colour difference information point and blood has reasonably been determined
The weight of pipe bending point, carries out curve fitting in conjunction with weight, so that even if having a small number of colour difference information points or vascular bending point inaccurate
Really, it still is able to obtain accurate optic cup line.The accuracy for improving optic cup line fitting result reduces artificial progress optic cup
The amount of labour of segmentation further improves the accuracy of diagnosis of glaucoma result.
Finally, it should be noted that the above embodiments are merely illustrative of the technical solutions of the present invention, rather than its limitations；Although
Present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that: it still may be used
To modify the technical solutions described in the foregoing embodiments or equivalent replacement of some of the technical features；
And these are modified or replaceed, technical solution of various embodiments of the present invention that it does not separate the essence of the corresponding technical solution spirit and
Range.
Claims (6)
1. the dividing method of optic cup in a kind of eyeground photograph characterized by comprising
The optic disk line in the photograph of eyeground is obtained, multiple colour difference information points in optic disk line region are obtained according to preset rules；
The multiple colour difference information point is carried out curve fitting, obtains initial optic cup line, and be arranged according to the initial optic cup line
Region of search；
Multiple vascular bending points in described search region are obtained, the multiple colour difference information point and the multiple blood vessel is curved
Song point, which carries out curve fitting, obtains optic cup line；
Wherein, the multiple colour difference information points obtained according to preset rules in optic disk line region, comprising:
The center for obtaining the corresponding region of the optic disk line is made a ray every predetermined angle using the center as endpoint, is obtained
Obtain the ray of respective numbers；
A pixel is taken every presetted pixel on each ray, is calculated corresponding along each pixel described in the directions of rays
Inside and outside value of chromatism, take the maximum aberration on the ray to be worth corresponding pixel as the color on the ray
Poor information point, wherein the pixel is in optic disk line region；
The step of carrying out curve fitting, obtain initial optic cup line the multiple colour difference information point specifically includes:
The rectangular coordinate system of the multiple colour difference information point is converted into polar coordinate system；
The multiple colour difference information point under polar coordinate system is fitted based on regularization least square approximating method, obtains pole
Curve under coordinate system；
The polar coordinate system of the curve is converted into rectangular coordinate system, using the curve under direct coordinate system as initial optic cup
Line；
The multiple colour difference information point and the multiple vascular bending point are carried out curve fitting the step of obtaining optic cup line tool
Body includes:
The rectangular coordinate system of the multiple colour difference information point and the multiple vascular bending point is converted into polar coordinate system；
It will be under the multiple colour difference information point and polar coordinate system under polar coordinate system based on regularization least square approximating method
The multiple vascular bending point is fitted as a whole, obtains the final curves under polar coordinate system；
The polar coordinate system of the final curves is converted into rectangular coordinate system, using the final curves under direct coordinate system as
Optic cup line；
The multiple colour difference information point and the multiple vascular bending point are carried out curve fitting and obtains optic cup line and includes:
All institutes where calculating the corresponding maximum aberration value of the colour difference information point and the colour difference information point on ray
The ratio for stating the color difference average value of pixel using the ratio as corresponding first reliability of the colour difference information point, and is counted
Calculate corresponding multiple second reliabilitys of the multiple vascular bending point；
It, can according to described second according to corresponding multiple first weights of the multiple colour difference information point of first reliability calculating
Corresponding multiple second weights of the multiple vascular bending point are calculated by degree；
According to the multiple first weight and the multiple second weight by the multiple colour difference information point and the multiple blood
Pipe bending point, which carries out curve fitting, obtains optic cup line；
It, can according to described second according to corresponding multiple first weights of the multiple colour difference information point of first reliability calculating
The step of calculating the multiple vascular bending point corresponding multiple second weights by degree specifically includes:
Step 1, regard all colour difference information points as first information point, regard all vascular bending points as second information
Point distributes to the first information point and second information point to by total weight 1 by following steps:
Using the average value of corresponding all first reliabilitys of all colour difference information points as color difference average of relatives value；
If the color difference average of relatives value is more than or equal to the first numerical value, enabling the weight of the first information point is the first ratio,
The weight of second information point is the second ratio；
If the color difference average of relatives value is less than or equal to second value, enabling the weight of the first information point is third ratio,
The weight of second information point is the 4th ratio；
If the color difference average of relatives value between first numerical value and the second value, enables the first information point
Weight be the 5th ratio, the weight of second information point is the 6th ratio；
Wherein, first numerical value is greater than the second value, first ratio and second ratio and be 1, it is described
Third ratio and the 4th ratio and be 1, the 5th ratio and the 6th ratio and be 1；
Step 2, the weight distribution of the first information point is given to each described colour difference information point by following formula, it will be described
The weight distribution of second information point gives each described vascular bending point:
First weight of any colour difference information point=first information point weight * (the first reliability of the colour difference information point/all
The sum of the first reliability of colour difference information point)；
The second weight=second information point weight * (the second reliability of the vascular bending point/all of any vascular bending point
The sum of the second reliability of vascular bending point)；
According to the multiple first weight and the multiple second weight by the multiple colour difference information point and the multiple blood
Pipe bending point the step of the obtaining optic cup line that carry out curve fitting specifically includes:
Each described first weight and each described second weight are multiplied by default value, using multiplied result as corresponding institute
State the number that colour difference information point and the corresponding vascular bending point occur；
According to number and coordinate value that each colour difference information point occurs, carry out curve fitting to all colour difference information points,
It according to the trip number and coordinate value of each vascular bending point, carries out curve fitting, obtains to all vascular bending points
Optic cup line.
2. the method according to claim 1, wherein described wrap according to the initial optic cup line setting region of search
It includes:
The inside curve and outside curve of the initial optic cup line, the inside curve are respectively set according to first reliability
Region between the outside curve is described search region.
3. the method according to claim 1, wherein the multiple vascular bending point of calculating is corresponding multiple
Second reliability includes: to obtain the corresponding vascular bending curvature of each vascular bending point and blood vessel diameter, the vascular bending
Curvature and the product of the corresponding blood vessel diameter are the second reliability of the corresponding vascular bending point.
4. the segmenting system of optic cup in a kind of eyeground photograph characterized by comprising
Module is obtained, for obtaining the optic disk line in the photograph of eyeground, is obtained according to preset rules multiple in optic disk line region
Colour difference information point；
Processing module obtains initial optic cup line, and according to described first for the multiple colour difference information point to carry out curve fitting
Region of search is arranged in beginning optic cup line；
Fitting module, for obtaining multiple vascular bending points in described search region, by the multiple colour difference information point and
The multiple vascular bending point, which carries out curve fitting, obtains optic cup line；
The acquisition module further includes ray acquisition submodule and Colorimetry submodule, in which:
Ray acquisition submodule is endpoint every pre- using the center for obtaining the center in the corresponding region of the optic disk line
If angle makees a ray, the ray of respective numbers is obtained；
Colorimetry submodule is calculated for taking a pixel every presetted pixel on each ray along the ray
Value of chromatism inside and outside each pixel described in direction is corresponding takes the corresponding pixel of the maximum aberration value on the ray to make
For the colour difference information point on the ray, wherein the pixel is in optic disk line region；
The processing module is specifically used for: the rectangular coordinate system of the multiple colour difference information point is converted to polar coordinate system；
The multiple colour difference information point under polar coordinate system is fitted based on regularization least square approximating method, obtains pole
Curve under coordinate system；
The polar coordinate system of the curve is converted into rectangular coordinate system, using the curve under direct coordinate system as initial optic cup
Line；
The fitting module is specifically used for: by the rectangular coordinate system of the multiple colour difference information point and the multiple vascular bending point
Be converted to polar coordinate system；
It will be under the multiple colour difference information point and polar coordinate system under polar coordinate system based on regularization least square approximating method
The multiple vascular bending point is fitted as a whole, obtains the final curves under polar coordinate system；
The polar coordinate system of the final curves is converted into rectangular coordinate system, using the final curves under direct coordinate system as
Optic cup line；
The processing module further includes the first reliability calculating submodule, in which: the first reliability calculating submodule, for calculating
All pixels where the corresponding maximum aberration value of colour difference information point and the colour difference information point on ray
Color difference average value ratio, using the ratio as corresponding first reliability of the colour difference information point；
The fitting module further includes the second reliability calculating submodule, weight calculation submodule and curve matching submodule,
In:
Second reliability calculating submodule, for calculating corresponding multiple second reliabilitys of the multiple vascular bending point；
Weight calculation submodule, for according to the multiple colour difference information point corresponding multiple first of first reliability calculating
Weight, according to corresponding multiple second weights of the multiple vascular bending point of second reliability calculating；
Curve matching submodule, for being believed the multiple color difference according to the multiple first weight and the multiple second weight
Breath point and the multiple vascular bending point carry out curve fitting and obtain optic cup line；
The weight calculation submodule is specifically used for:
Step 1, regard all colour difference information points as first information point, regard all vascular bending points as second information
Point distributes to the first information point and second information point to by total weight 1 by following steps:
Using the average value of corresponding all first reliabilitys of all colour difference information points as color difference average of relatives value；
If the color difference average of relatives value is more than or equal to the first numerical value, enabling the weight of the first information point is the first ratio,
The weight of second information point is the second ratio；
If the color difference average of relatives value is less than or equal to second value, enabling the weight of the first information point is third ratio,
The weight of second information point is the 4th ratio；
If the color difference average of relatives value between first numerical value and the second value, enables the first information point
Weight be the 5th ratio, the weight of second information point is the 6th ratio；
Wherein, first numerical value is greater than the second value, first ratio and second ratio and be 1, it is described
Third ratio and the 4th ratio and be 1, the 5th ratio and the 6th ratio and be 1；
Step 2, the weight distribution of the first information point is given to each described colour difference information point by following formula, it will be described
The weight distribution of second information point gives each described vascular bending point:
First weight of any colour difference information point=first information point weight * (the first reliability of the colour difference information point/all
The sum of the first reliability of colour difference information point)；
The second weight=second information point weight * (the second reliability of the vascular bending point/all of any vascular bending point
The sum of the second reliability of vascular bending point)；
The curve matching submodule is specifically used for:
Each described first weight and each described second weight are multiplied by default value, using multiplied result as corresponding institute
State the number that colour difference information point and the corresponding vascular bending point occur；
According to number and coordinate value that each colour difference information point occurs, carry out curve fitting to all colour difference information points,
It according to the trip number and coordinate value of each vascular bending point, carries out curve fitting, obtains to all vascular bending points
Optic cup line.
5. system according to claim 4, which is characterized in that the processing module further includes that region of search determines submodule
Block, in which:
Region of search determines submodule, for the inside curve of the initial optic cup line to be respectively set according to first reliability
With outside curve, the region between the inside curve and the outside curve is described search region.
6. system according to claim 4, which is characterized in that the second reliability calculating submodule is specifically used for: obtaining institute
The corresponding vascular bending curvature of each vascular bending point and blood vessel diameter are stated, the vascular bending curvature and the corresponding blood vessel are straight
The product of diameter is the second reliability of the corresponding vascular bending point.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610810816.XA CN106446805B (en) | 2016-09-08 | 2016-09-08 | A kind of eyeground shine in optic cup dividing method and system |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201610810816.XA CN106446805B (en) | 2016-09-08 | 2016-09-08 | A kind of eyeground shine in optic cup dividing method and system |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN106446805A CN106446805A (en) | 2017-02-22 |
| CN106446805B true CN106446805B (en) | 2019-09-13 |
Family
ID=58164537
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201610810816.XA Active CN106446805B (en) | 2016-09-08 | 2016-09-08 | A kind of eyeground shine in optic cup dividing method and system |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN106446805B (en) |
Families Citing this family (5)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN110037651B (en) * | 2018-01-15 | 2022-03-25 | 江威 | Method and device for controlling quality of fundus image |
| CN109166095B (en) * | 2018-07-11 | 2021-06-25 | 广东技术师范学院 | Fundus image cup and disc segmentation method based on generation countermeasure mechanism |
| CN110969617B (en) * | 2019-12-17 | 2024-03-15 | 腾讯医疗健康(深圳)有限公司 | Method, device, equipment and storage medium for identifying video cup video disc image |
| CN111863241B (en) * | 2020-07-10 | 2023-06-30 | 北京化工大学 | Fundus imaging classification system based on integrated deep learning |
| CN114387210A (en) * | 2021-12-03 | 2022-04-22 | 依未科技(北京)有限公司 | Method, apparatus, medium, and device for fundus feature acquisition |
Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102112044A (en) * | 2008-05-14 | 2011-06-29 | 科学、技术与研究机构 | Automatic cup-to-disc ratio measurement system |
| CN102316789A (en) * | 2008-12-15 | 2012-01-11 | 新加坡保健服务集团有限公司 | Confirm the method and system of the position of cup boundary |
| JP5007420B2 (en) * | 2006-09-21 | 2012-08-22 | タック株式会社 | Image analysis system and image analysis program |
-
2016
- 2016-09-08 CN CN201610810816.XA patent/CN106446805B/en active Active
Patent Citations (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| JP5007420B2 (en) * | 2006-09-21 | 2012-08-22 | タック株式会社 | Image analysis system and image analysis program |
| CN102112044A (en) * | 2008-05-14 | 2011-06-29 | 科学、技术与研究机构 | Automatic cup-to-disc ratio measurement system |
| CN102316789A (en) * | 2008-12-15 | 2012-01-11 | 新加坡保健服务集团有限公司 | Confirm the method and system of the position of cup boundary |
Non-Patent Citations (3)
| Title |
|---|
| 《Accurate and Efficient Optic Disc Detection and Segmentation by a Circular Transformation》;Lu Shijian;《IEEE Transactions on Medical Imaging》;20110812;第30卷(第12期);正文第3部分 * |
| 《Optic Disk and Cup Segmentation from Monocular Colour Retinal Images for Glaucoma Assessment》;Joshi GD et al;《IEEE Transactions on Medical Imaging》;20110502;第30卷(第6期);正文第3部分粗选择 * |
| 《基于眼底图像的青光眼形态特征自动分析的研究》;吴晗辉;《中国优秀硕士学位论文全文数据库医药卫生科技辑》;20150515(第2015年第05期);全文 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN106446805A (en) | 2017-02-22 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN106446805B (en) | A kind of eyeground shine in optic cup dividing method and system | |
| Wong et al. | Level-set based automatic cup-to-disc ratio determination using retinal fundus images in ARGALI | |
| CN109166124A (en) | A kind of retinal vascular morphologies quantization method based on connected region | |
| CN110555845A (en) | Fundus OCT image identification method and equipment | |
| CN107563996B (en) | A kind of novel discus nervi optici dividing method and system | |
| JP2019192215A (en) | 3d quantitative analysis of retinal layers with deep learning | |
| WO2021190656A1 (en) | Method and apparatus for localizing center of macula in fundus image, server, and storage medium | |
| CN108618749A (en) | Retinal vessel three-dimensional rebuilding method based on portable digital fundus camera | |
| CN102567734B (en) | Specific value based retina thin blood vessel segmentation method | |
| CN108510493A (en) | Boundary alignment method, storage medium and the terminal of target object in medical image | |
| Zhang et al. | Convex hull based neuro-retinal optic cup ellipse optimization in glaucoma diagnosis | |
| CN106846293A (en) | Image processing method and device | |
| CN113768461A (en) | Fundus image analysis method and system and electronic equipment | |
| CN110288574A (en) | A kind of adjuvant Ultrasonographic Diagnosis hepatoncus system and method | |
| Choudhary et al. | ANN glaucoma detection using cup-to-disk ratio and neuroretinal rim | |
| CN108665474A (en) | A kind of eye fundus image Segmentation Method of Retinal Blood Vessels based on B-COSFIRE | |
| CN116309235A (en) | Fundus image processing method and system for diabetes prediction | |
| CN111784641B (en) | Neural image curvature estimation method and device based on topological structure | |
| CN110598652B (en) | Fundus data prediction method and device | |
| Tan et al. | Automatic feature extraction of optical coherence tomography for lamina cribrosa detection | |
| CN116407080A (en) | Evolution identification and 3D visualization system and method for fundus structure of myopic patient | |
| Thanh et al. | A real-time classification of glaucoma from retinal fundus images using AI technology | |
| CN115762787A (en) | Eyelid disease surgery curative effect evaluation method and system based on eyelid topological morphology analysis | |
| Suryawanshi | An approach to glaucoma using image segmentation techniques | |
| Zhou et al. | Computer aided diagnosis for diabetic retinopathy based on fundus image |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| C06 | Publication | ||
| PB01 | Publication | ||
| C10 | Entry into substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant | ||
| TR01 | Transfer of patent right | ||
| TR01 | Transfer of patent right |
Effective date of registration: 20231211 Address after: 215000 Bamboo Garden Road, Suzhou high tech Zone, Jiangsu Province, No. 209 Patentee after: Suzhou Zichuan Information Technology Co.,Ltd. Address before: 100029, No. 15 East Third Ring Road, Chaoyang District, Beijing Patentee before: BEIJING University OF CHEMICAL TECHNOLOGY |