CN119359732B - A method, device and medium for analyzing red eye grade combined with vascular morphology analysis - Google Patents
A method, device and medium for analyzing red eye grade combined with vascular morphology analysis Download PDFInfo
- Publication number
- CN119359732B CN119359732B CN202411937407.7A CN202411937407A CN119359732B CN 119359732 B CN119359732 B CN 119359732B CN 202411937407 A CN202411937407 A CN 202411937407A CN 119359732 B CN119359732 B CN 119359732B
- Authority
- CN
- China
- Prior art keywords
- blood vessel
- area
- region
- red
- vascular
- 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
Landscapes
- Image Analysis (AREA)
Abstract
The application provides an eye red grade analysis method, equipment and medium combining vascular morphology analysis, and relates to the technical field of image processing, wherein the method comprises the steps of inputting a target eye image into a pre-trained segmentation model to obtain a conjunctiva area and a cornea area, determining a limbal area based on the conjunctiva area and the cornea area, extracting red pixel areas in the conjunctiva area and the limbal area, performing skeletonizing treatment to obtain a skeletonized mask image, expanding the skeletonized mask image, and performing exclusive OR operation with the red pixel areas to obtain a bleeding area; the blood vessel region in the red pixel region is determined based on the bleeding region and the blood vessel morphology value is measured, the ocular red grade is determined based on the area ratio of the red pixel region in the conjunctival region and the limbus region, and the ocular red grade is further corrected by the blood vessel morphology value. The application can accurately separate blood vessels and combine the blood vessel morphology to determine the eye red grade, and the eye red analysis is more comprehensive and scientific.
Description
Technical Field
The invention relates to the technical field of ophthalmic image processing, in particular to an eye red grade analysis method, equipment and medium combined with vascular morphology analysis.
Background
Traditional eye red analysis depends on visual observation of doctors and judges the severity according to the experience of the doctors, however, the visual observation often has strong subjectivity and randomness, and the accuracy is not high. Conventional image processing techniques analyze the red area fraction by thresholding to determine the eye red level. However, the background color of the eye surface image is not constant due to the technical effects of illumination, device chromatic aberration and photographer, and the analysis of the red ratio by color extraction alone often has high false recognition rate and low eye red analysis precision.
A method, equipment and medium for determining eye red grade are disclosed in China patent document with the application number of CN202311743538.7, and the method comprises the steps of obtaining an eye red area image by dividing a target eye image with congestion of an eye surface, carrying out blood vessel division on the eye red area image to determine a blood vessel occupation ratio in the eye red area image, determining an eye red index based on the eye red area image, and obtaining the eye red grade of the target eye image through an eye red grade classification model based on the blood vessel occupation ratio and the eye red index. According to the method, the ocular red index and the blood vessel occupation ratio can be combined to carry out ocular red grade grading, and the accuracy of ocular red grade grading can be improved.
Disclosure of Invention
The embodiment of the application provides an eye red grade analysis method, an eye red grade analysis device and an eye red grade analysis medium combined with blood vessel morphology analysis, which are used for solving the technical problems of single eye red analysis grade index and poor eye red grade accuracy.
The following describes a specific technical scheme provided by the embodiment of the application.
In a first aspect, there is provided a method of eye red grade analysis in combination with vascular morphology analysis, comprising:
Inputting a target eye image into a pre-trained segmentation model for segmentation to obtain a conjunctiva area and a cornea area, and determining a limbus area based on the conjunctiva area and the cornea area;
extracting red pixel areas in the conjunctiva area and the limbus area, performing skeletonizing treatment to obtain a skeletonized mask image, expanding the skeletonized mask image, and performing exclusive-or operation with the red pixel areas to obtain a bleeding area;
determining a blood vessel region of the red pixel region other than the bleeding region based on the bleeding region, and measuring a blood vessel morphology value of the blood vessel region, the blood vessel morphology value including one or more of a blood vessel width, a blood vessel topology, and a blood vessel density;
And determining an eye red grade based on the area of the red pixel region, and further correcting the eye red grade through the blood vessel morphology value.
In a possible embodiment, the eye red level analysis method combined with the blood vessel morphology analysis further includes:
And determining that the width of the blood vessel is larger than a first preset threshold value, or the topological structure of the blood vessel is a netlike blood vessel and the density of the blood vessel is larger than a second preset threshold value, and improving the level of redness.
In one possible embodiment, the step of measuring the width of the blood vessel specifically comprises:
carrying out refinement treatment on the vessel tree mask to obtain the central line of each vessel;
respectively calculating the distance from the point on the blood vessel central line to the nearest background and determining the distance as the blood vessel radius;
the vessel radius is multiplied by two to obtain the vessel width.
In one possible embodiment, measuring the vessel topology specifically comprises:
detecting a key point in the blood vessel region, wherein the key point comprises an endpoint of only one neighbor pixel, a branching point of two neighbor pixels and a crossing point of at least three neighbor pixels;
tracking from each end point along the vessel centerline to the next branch point or intersection, recording path length and direction changes;
And calculating the connectivity of the blood vessel, the number of branch points, the length of the blood vessel and the angle of the blood vessel branches to obtain the blood vessel topological structure.
In one possible embodiment, measuring the blood vessel density specifically includes:
Calculating the total number of all non-zero pixels on the vascular tree mask to obtain a first numerical value;
dividing the first value by the total area of the vessel tree mask to obtain the vessel density.
In one possible embodiment, measuring the blood vessel density specifically includes:
Carrying out refinement treatment on the vascular tree mask;
traversing the thinned image, and calculating the total number of all non-zero pixels to obtain a second value;
dividing the second value by the total area of the vascular tree mask to obtain the vascular density.
In one possible embodiment, after extracting the red pixel regions within the conjunctival region and the limbus region, prior to skeletonizing the red pixel regions, further comprising complementing the red pixel regions using a region growing algorithm.
In a second aspect, there is provided an eye red grade analysis apparatus in combination with a vessel morphology analysis, comprising:
A memory for storing program instructions;
A processor, configured to invoke the program instructions stored in the memory to implement the eye red level analysis method according to any one of the first aspect in combination with vessel morphology analysis.
In a third aspect, a computer readable storage medium is provided, where the computer readable storage medium stores a program code for implementing the eye red level analysis method according to any one of the first aspects in combination with vessel morphology analysis.
The eye red grade analysis method has the advantages that the blood vessels are accurately separated and measured by the steps of skeletonizing, expanding, exclusive-or operation and the like on the red pixel area (eye red area), and the eye red grade is determined by combining the blood vessel morphology and the eye red area ratio. The method not only analyzes the ocular red area ratio and the vascular area ratio, but also fully considers the vascular morphological characteristics such as the vascular width, the vascular topological structure, the vascular density and the like, and the ocular red analysis is more comprehensive and scientific.
Drawings
Fig. 1 is a flowchart of an eye red level analysis method combined with blood vessel morphology analysis according to an embodiment of the present application.
Fig. 2 is a flowchart of the implementation of step S101.
Detailed Description
In order that the objects, features and advantages of the application will be more readily understood, a more particular description of the application will be rendered by reference to specific embodiments thereof which are illustrated in the appended drawings. In the following description, numerous specific details are set forth in order to provide a thorough understanding of the present application, but the present application may be practiced in other ways than as described herein, and therefore the present application is not limited to the specific embodiments disclosed below.
The embodiment of the application is applied to image analysis of the eye red image, and can be used for carrying out blood vessel morphology measurement after dividing the eye congestion image and separating blood vessel areas in the eye red, and determining the eye red grade by combining blood vessel morphology values, so that the eye red disease analysis is more detailed and comprehensive.
The embodiments of the present application will be described in further detail below with reference to the drawings in the specification and by way of specific embodiments, and it should be understood that the embodiments described herein are for illustration and explanation of the present application, and are not intended to limit the present application, and the features of the embodiments and embodiments of the present application may be combined with each other without conflict.
Before describing an eye-red level analysis method combined with vessel morphology analysis provided by the embodiment of the present application, for convenience of understanding, a technical background of the embodiment of the present application will be described in detail. In addition to ocular congestion, which is a condition in which blood vessels (more commonly, single thicker blood vessels, or multiple finer blood vessels criss-cross to form a net), are present, there is a condition in which blood vessels are uniformly and uniformly reddened, and blood vessels are vague (diffuse), and such bleeding is clinically often indicative of severe inflammation or lasting in duration, and may even be accompanied by infiltration, edema, proliferation or degeneration of tissues. The analysis of the eye red in the prior art is often to convert the image into HSV color mode or LAB color mode, identifying red pixels in the conjunctiva area by reading the H-channel in HSV, or A, B-channel in LAB. However, the eye surface image is exposed to light, the device has a color difference, and the photographer has a large influence, so that the effects of uniform brightness and fixed color are difficult to achieve, and the traditional scheme has a high false recognition rate (such as detecting the background as a blood vessel or not detecting the blood vessel). In the prior art, a method, a device and a medium for determining an eye red grade are disclosed in China patent No. CN202311743538.7, which discloses a method for determining an eye red grade, comprising the steps of segmenting a target eye image with congestion of an eye surface to obtain an eye red area image, segmenting blood vessels in the eye red area image to determine a blood vessel occupation ratio in the eye red area image, determining an eye red index based on the eye red area image, and obtaining an eye red grade of the target eye image through an eye red grade classification model based on the blood vessel occupation ratio and the eye red index, wherein the eye red grade classification model is trained through a training sample set comprising blood vessel occupation ratio characteristics, eye red index characteristics and corresponding eye red grade labels. However, the method still only considers the area occupation ratio of the blood vessels, and the evaluation index is still single and still needs to be improved.
As shown in FIG. 1, an embodiment of the present application provides an eye red level analysis method combined with blood vessel morphology analysis. The method comprises the following steps:
step S101, inputting a target eye image into a pre-trained segmentation model for segmentation to obtain a conjunctiva area and a cornea area, and determining a limbal area based on the conjunctiva area and the cornea area;
In some specific implementations of this step, the eye with the ocular surface congestion is photographed to obtain a target eye image to be detected of the ocular surface congestion, where the target eye image contains an eye red area caused by vasodilation and/or diffuse hemorrhage. In some embodiments, referring to fig. 2, step S101 may specifically include the steps of:
step S1011, obtaining a target eye diagram and carrying out normalization pretreatment such as brightness, saturation and the like;
Step S1012, inputting the preprocessed image data into a pre-trained neural network segmentation model, wherein the model can output segmented conjunctiva areas and cornea areas by using GAN, U-NET, YOLO and the like;
step S1013, obtaining a limbal region by image dilation and image exclusive or operation of the corneal region and the corneal region.
Prior to step S1013, step S101 may further include obtaining an accurate conjunctiva region by edge fitting to improve accuracy of segmentation.
Step S102, extracting red pixel areas in the conjunctiva area and the limbus area, performing skeletonizing treatment to obtain a skeletonized mask image, expanding the skeletonized mask image, and performing exclusive OR operation with the red pixel areas to obtain a bleeding area;
in some implementations of this step, the means for extracting red pixel regions within the conjunctiva region and the limbus region may include:
The first way is to convert the image into HSV color mode or LAB color mode, and to identify red pixels in conjunctiva area to obtain red pixel area by reading H channel in HSV or A, B channel in LAB.
In the second mode, the image is converted into a gray color mode, and a blood vessel region is obtained through sobel/scharr edge detection and double-threshold connection, so that a red pixel region is obtained.
And in a third mode, converting the image into a gray color mode, and acquiring a blood vessel region by a blood vessel extraction method based on morphology to obtain a red pixel region.
In a fourth mode, a blood vessel region is obtained based on a blood vessel extraction method of an active contour model (a snake line model), and a red pixel region is obtained.
In some embodiments of the present application, enhancing the blood vessel region by a Frangi filtering algorithm based on a Hessian matrix may also be included before extracting the red pixel region. The red pixel region extraction is easier, more comprehensive and more accurate after the enhancement.
Step S103 of determining a blood vessel region other than the bleeding region in the red pixel region based on the bleeding region, and measuring a blood vessel morphology value of the blood vessel region, the blood vessel morphology value including one or more of a blood vessel width, a blood vessel topology, and a blood vessel density;
through step S102 and step S103, the blood vessel region in the red pixel region can be separated.
In some implementations of step S103, measuring the vessel width of the vessel region may include:
Calculating the width of the blood vessel according to the edge coordinates of the blood vessel, and determining the edge coordinates of the blood vessel can comprise the steps of marking the blood vessels into a1 st blood vessel, a2 nd blood vessel, a.once the marking is finished, and obtaining the edge coordinates of the blood vessel.
In other embodiments of step S103, measuring the vessel width of the vessel region may include:
The vessel mask is refined to obtain vessel center lines, distances from points on the vessel center lines to the nearest background are calculated respectively, the distances are determined to be vessel radius, and the vessel radius is multiplied by 2 to obtain the vessel width. The specific implementation conditions are as follows:
the vessel tree mask is refined to obtain the center line of each vessel. This can be achieved by morphological operations such as erosion until all vascular lines become as thin as possible.
And applying distance transformation to points on the central line of the vascular mask to obtain the distance from each point to the nearest background point.
For each point on the centerline after refinement, the smallest distance value, i.e. the vessel radius at that location, can be found around it.
By calculating twice the distance value at these points, the vessel width can be obtained.
Measuring the width of the blood vessel helps to judge the proliferation of the blood vessel and the inflammatory response. For patients receiving treatment, tracking changes in vascular width can be used to assess the efficacy of the treatment or the progression of the condition.
In some implementations of step S103, measuring the vessel topology may specifically include:
detecting a key point in the blood vessel region, wherein the key point comprises an endpoint of only one neighbor pixel, a branching point of two neighbor pixels and a crossing point of at least three neighbor pixels;
Starting from each end point, tracking the next branch point or crossing point along the central line of the blood vessel, recording the path length and direction change, and calculating the connectivity of the blood vessel, the number of branch points, the length of the blood vessel and the angle of the branch of the blood vessel to obtain the topological structure of the blood vessel. The specific embodiment for measuring the vascular topology is as follows:
First, detect keypoints, including (endpoint: only one neighbor, branch point: two neighbor, cross point: three or more neighbor.)
Second, from each end point, the next branch point or intersection is traced along the vessel centerline, and information such as path length and direction change is recorded.
Thirdly, calculating connectivity of the blood vessel, and determining whether the blood vessel network is completely connected.
Fourth, branch mode, analyzing the number and distribution of branch points.
Fifth, the length of the blood vessel is calculated.
And sixthly, angle analysis, namely calculating the angle of the blood vessel branch, and relieving the bending degree of the blood vessel.
Knowledge of the vascular topology in normal and pathological conditions helps reveal the mechanism of disease progression. Certain ocular surface diseases can result in abnormal changes in the vascular network, such as vasodilation in dry eye, neovascularization in diabetic retinopathy. Diagnosis of these diseases can be aided by analysis of the topology of the blood vessels.
In some embodiments of step S103, measuring the blood vessel density may specifically include:
Calculating the area ratio of the vascular pixels in the conjunctiva area and the limbus area to obtain a first density;
Calculating the length of blood vessels per unit area in the conjunctival region and in the limbal region to obtain a second density,
A vessel density is derived based on the first density and/or the second density.
Specifically, the blood vessel density is divided into two types, a blood vessel pixel density and a blood vessel length density. Wherein the vessel pixel density is defined as the ratio of the number of vessel pixels to the total area, and the calculating step comprises:
first, the total number of all non-zero pixels on the vessel tree mask is calculated.
Second, dividing the value by the total area of the mask.
Wherein the vessel length density is defined as the vessel length per unit area, the calculating step comprising:
And firstly, carrying out thinning treatment on the vascular tree mask.
And secondly, traversing the thinned image, and calculating the total number of all non-zero pixels.
Third, dividing the value by the total area of the mask.
The change of ocular surface vascular density can be used as symptom index of various diseases, such as diabetic retinopathy, hypertensive retinopathy, abnormal vascular distribution of xerophthalmia, etc. These diseases can lead to abnormal proliferation or atrophy of blood vessels, thereby altering vascular density. For patients who have been diagnosed, periodic monitoring of changes in vascular density is helpful in assessing the efficacy of treatment and the progression of the disease.
Step S104, determining an eye red grade based on the area ratio of the red pixel area in the conjunctiva area and the limbal area, and further correcting the eye red grade through the vascular morphology value.
In one possible embodiment, the eye red level analysis method combined with the blood vessel morphology analysis further includes:
And determining that the width of the blood vessel is larger than a first preset threshold, the topological structure of the blood vessel is a netlike blood vessel, and the density of the blood vessel is larger than a second preset threshold, and then improving the level of redness.
In one possible embodiment, the eye red grade analysis method combined with the blood vessel morphology analysis further comprises determining that the bleeding area is greater than a third preset threshold, and increasing the eye red grade.
In one possible embodiment, the eye red level analysis method combined with the blood vessel morphology analysis further comprises the steps of determining that the brightness information of the red pixel area is larger than a fourth preset threshold value, and then increasing the eye red level, or determining that the saturation information of the red pixel area is larger than a fifth preset threshold value, and then increasing the eye red level.
For ease of understanding, the flow of determining the level of redness is illustrated below. Conventional eye red grade classification is generally classified into five grades of 0 to 4, for example, eye red grade of 1 for conjunctival congestion is determined by calculating the area ratio of red pixel area in conjunctival area, and eye red grade of 2 for ciliary congestion is determined by calculating the area ratio of eye red area in limbal area.
If the vessel width is determined to be greater than or equal to the preset vessel width threshold, the eye red level is increased (e.g., 0.5 level is added to the eye red level determined based on the area ratio, in this example, the eye red level is determined to be (1+2)/2+0.5=2).
In yet another embodiment of the present application, the method further comprises, after extracting the red pixel regions within the conjunctival region and the limbus region, complementing the red pixel regions with a region growing algorithm prior to skeletonizing the red pixel regions.
In some implementations of embodiments of the application, if there are multiple distinct regions in the image that need to be segmented, the above process may be repeated, selecting new seed points for each region of interest and performing region growing.
The eye red grade analysis method combined with the blood vessel morphology analysis can accurately separate blood vessels in the eye red, measure the blood vessel morphology, and judge the eye red grade by combining with the blood vessel morphology value, so that the eye red analysis is more comprehensive and scientific.
In order to remove the interference and improve the accuracy of the eye red analysis, in one possible embodiment, before the extracting the red pixel regions in the conjunctiva region and the limbus region, the method further comprises removing the eyelash shielding region and the pigmentation region in the conjunctiva region and the limbus region by screening the brightness value and/or the saturation value.
In one possible embodiment, the eye red grade analysis method combined with vascular morphology analysis further comprises the steps of acquiring a target eye image of eye congestion and preprocessing, wherein the preprocessing comprises brightness normalization and/or saturation normalization.
In one possible embodiment, the eye red level analysis method combined with the blood vessel morphology analysis further comprises the steps of constructing and training a segmentation model for segmenting conjunctival regions and cornea regions, wherein the segmentation model adopts a network structure of GAN, U-NET or YOLO.
A further embodiment of the present application provides an eye red level analysis apparatus in combination with a blood vessel morphology analysis, including:
A memory for storing program instructions;
a processor for invoking the program instructions stored in the memory to implement the eye red level analysis method in combination with vessel morphology analysis as described in any one of the embodiments above.
In yet another embodiment of the present application, there is also provided a computer readable storage medium storing a program code for implementing the eye red level analysis method in combination with vessel morphology analysis according to any one of the above embodiments.
All or part of the steps in the various methods of the above embodiments may be performed by controlling related hardware by a program, which may be stored in a readable storage medium including Read-Only Memory (ROM), random access Memory (Random Access Memory, RAM), programmable Read-Only Memory (Programmable Read-Only Memory, PROM), erasable programmable Read-Only Memory (ErasableProgrammable Read Only Memory, EPROM), one-time programmable Read-Only Memory (One-timeProgrammable Read-Only Memory, OTPROM), electrically erasable rewritable Read-Only Memory (EEPROM), compact disc Read-Only Memory (CD-ROM) or other optical disc Memory, magnetic disk Memory, tape Memory, or any other medium capable of being used for carrying or storing data.
The above description is only of the preferred embodiments of the present invention and is not intended to limit the present invention, but various modifications and variations can be made to the present invention by those skilled in the art. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention should be included in the protection scope of the present invention.
Claims (9)
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411937407.7A CN119359732B (en) | 2024-12-26 | 2024-12-26 | A method, device and medium for analyzing red eye grade combined with vascular morphology analysis |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202411937407.7A CN119359732B (en) | 2024-12-26 | 2024-12-26 | A method, device and medium for analyzing red eye grade combined with vascular morphology analysis |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN119359732A CN119359732A (en) | 2025-01-24 |
| CN119359732B true CN119359732B (en) | 2025-03-25 |
Family
ID=94315109
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202411937407.7A Active CN119359732B (en) | 2024-12-26 | 2024-12-26 | A method, device and medium for analyzing red eye grade combined with vascular morphology analysis |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN119359732B (en) |
Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102843957A (en) * | 2009-08-24 | 2012-12-26 | 新加坡保健服务集团有限公司 | Method and system for detecting disc haemorrhages |
| CN117314847A (en) * | 2023-09-21 | 2023-12-29 | 上海美沃精密仪器股份有限公司 | Eye red state analysis method and device, computing equipment and storage medium |
Family Cites Families (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN103226689B (en) * | 2012-01-30 | 2016-04-13 | 展讯通信（上海）有限公司 | Detect blood-shot eye illness method and device, removal blood-shot eye illness method and device |
| CN117197064A (en) * | 2023-08-30 | 2023-12-08 | 河海大学 | Automatic non-contact eye red degree analysis method |
| CN117788891A (en) * | 2023-12-18 | 2024-03-29 | 杭州微晓医疗科技有限公司 | Method, equipment and medium for determining eye red grade |
-
2024
- 2024-12-26 CN CN202411937407.7A patent/CN119359732B/en active Active
Patent Citations (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN102843957A (en) * | 2009-08-24 | 2012-12-26 | 新加坡保健服务集团有限公司 | Method and system for detecting disc haemorrhages |
| CN117314847A (en) * | 2023-09-21 | 2023-12-29 | 上海美沃精密仪器股份有限公司 | Eye red state analysis method and device, computing equipment and storage medium |
Also Published As
| Publication number | Publication date |
|---|---|
| CN119359732A (en) | 2025-01-24 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Gandhi et al. | Diagnosis of diabetic retinopathy using morphological process and SVM classifier | |
| CN105009174B (en) | Method and system for assessing tissue fibrosis | |
| CN111340789A (en) | Method, device, equipment and storage medium for identifying and quantifying eye fundus retinal blood vessels | |
| CN118115466B (en) | A method for detecting pseudo-lesions of fundus | |
| CN114830173A (en) | Method for determining the severity of skin disorders based on the percentage of human body surface area covered by lesions | |
| CN112734828A (en) | Method, device, medium and equipment for determining center line of fundus blood vessel | |
| CN113344894A (en) | Method and device for extracting characteristics of eyeground leopard streak spots and determining characteristic index | |
| Mendonça et al. | Segmentation of the vascular network of the retina | |
| KR20180045473A (en) | System, method and computer program for melanoma detection using image analysis | |
| CN113344895A (en) | High-precision fundus blood vessel diameter measuring method, device, medium and equipment | |
| Mohammad et al. | Texture analysis for glaucoma classification | |
| Zhang et al. | Retinal spot lesion detection using adaptive multiscale morphological processing | |
| CN119359732B (en) | A method, device and medium for analyzing red eye grade combined with vascular morphology analysis | |
| CN118247301A (en) | Human body temperature region identification method and system based on infrared thermal image | |
| Guatemala-Sanchez et al. | Nuclei segmentation on histopathology images of breast carcinoma | |
| Wu et al. | Retinal vessel radius estimation and a vessel center line segmentation method based on ridge descriptors | |
| CN112381767B (en) | Cornea reflection image screening method and device, intelligent terminal and storage medium | |
| CN115100178A (en) | Method, device, medium and equipment for evaluating morphological characteristics of fundus blood vessels | |
| Roy et al. | A novel computer aided quantification method of focal arteriolar narrowing using colour retinal image | |
| Wisaeng | CFLHCF: Simultaneous detection of the optic disc and exudates using color features, local homogeneity and contextual features | |
| Welikala et al. | Differing matched filter responsivity for the detection of proliferative diabetic retinopathy | |
| CN115457038B (en) | Training method of hierarchical prediction model, hierarchical prediction method and related products | |
| CN119048531B (en) | Ear-nose-throat examination image segmentation method and system | |
| CN112734784B (en) | Method, device, medium and equipment for determining fundus blood vessel boundary with high precision | |
| Srivastava et al. | Automatic extraction of micro-aneurysms and haemorrhages from digital fundus image |
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 |