Intelligent enhancement method for fundus blood vessel image

Abstract

The invention relates to the technical field of fundus blood vessel image enhancement, in particular to an intelligent fundus blood vessel image enhancement method. The method comprises the steps of firstly obtaining a fundus blood vessel image and a blood vessel area; further obtaining a corrosion index of each blood vessel region according to the change characteristics of the number of the pixel points before and after corrosion; further according to the difference characteristic of the corrosion index of the blood vessel region in the preset neighborhood window of the target region, obtaining a fine blood vessel coefficient of the target region, and obtaining a potential fine blood vessel region; further, according to difference characteristics of the corrosion indexes of the potential fine blood vessel region and the corrosion indexes of the adjacent blood vessel region and distribution characteristics of all the corrosion indexes, in combination with fine blood vessel coefficients, obtaining an authenticity coefficient of the potential fine blood vessel region, and obtaining the fine blood vessel region; and finally, according to the corrosion index and the authenticity coefficient of the small blood vessel region, obtaining an enhancement coefficient of the small blood vessel region, and performing image enhancement on the fundus blood vessel image.

Description

Technical Field

[0001] The present invention relates to the technical field of fundus vascular image enhancement, and in particular to a fundus vascular image intelligent enhancement method. Background Art

[0002] As people's lifestyles change, various electronic products have penetrated into all aspects of life. Various unhealthy eye habits and excessive eye fatigue have caused more and more patients with eye diseases. Various eye diseases will cause different degrees of changes in the fundus structure. Fundus images are the most direct and effective basis for doctors to diagnose eye diseases. The retina at the bottom of the human eyeball contains a large number of tiny blood vessels. The changes in the trend, diameter, width, shape and structure of these microvessels are related to eye diseases, cardiovascular and cerebrovascular diseases, hypertension and other systemic diseases.

[0003] In the existing process of enhancing fundus vascular images, the vascular skeleton is extracted through morphological corrosion, the vascular breakpoints are obtained to determine the part with smaller vascular diameter, and the area with smaller vascular diameter is enhanced; however, when extracting the vascular skeleton, problems such as excessive reduction, inaccurate judgment of breakpoints, and noise interference may occur, resulting in inaccurate judgment of small blood vessels and affecting image quality; in addition, the macular area may also interfere with the judgment of small blood vessels and affect the enhancement effect. Summary of the invention

[0004] In order to solve the existing technical problems of inaccurate positioning of small blood vessel areas and unsatisfactory image enhancement effects, the purpose of the present invention is to provide an intelligent enhancement method for fundus vascular images. The technical solutions adopted are as follows: A method for intelligently enhancing fundus vascular images, the method comprising: Acquire fundus vascular images and vascular regions; perform morphological corrosion on each of the vascular regions; According to the change characteristics of the number of pixel points before and after corrosion, the corrosion index of each blood vessel area is obtained; from the blood vessel root to any end point is recorded as a blood vessel branch; any blood vessel branch is selected as a target branch; any blood vessel area on the target branch is selected as a target area; according to the difference characteristics of the corrosion index of the blood vessel area in the preset neighborhood window of the target area, the small blood vessel coefficient of the target area is obtained; according to the small blood vessel coefficient, the potential small blood vessel area is obtained; According to the difference characteristics between the corrosion index of the potential small blood vessel area and the corrosion index of the adjacent blood vessel area, and the distribution characteristics of all the corrosion indexes, combined with the small blood vessel coefficient, the authenticity coefficient of the potential small blood vessel area is obtained; according to the authenticity coefficient, the small blood vessel area is obtained; according to the corrosion index and the authenticity coefficient of each small blood vessel area, the enhancement coefficient of each small blood vessel area is obtained; The fundus blood vessel image is enhanced according to the enhancement coefficient.

[0005] Furthermore, the method for obtaining the corrosion index includes: In each of the blood vessel regions, the ratio of the change in the number of pixel points before and after corrosion to the number of pixel points before corrosion is used as a corrosion index for each of the blood vessel regions.

[0006] Furthermore, the method for obtaining the small blood vessel coefficient includes: The preset neighborhood window is centered on the target area; According to the difference characteristics between the corrosion index of the target area and the corrosion index of the adjacent blood vessel areas on both sides, combined with the maximum difference of the corrosion index between all the blood vessel areas in the preset neighborhood window, the small blood vessel coefficient of the target area is obtained; the difference characteristics between the corrosion index of the target area and the corrosion index of the adjacent blood vessel areas on both sides are positively correlated with the small blood vessel coefficient; the maximum difference is negatively correlated with the small blood vessel coefficient.

[0007] Furthermore, the method for obtaining the potential small blood vessel area includes: When the small blood vessel coefficient is greater than or equal to a first preset threshold, the corresponding blood vessel region is marked as a potential small blood vessel region.

[0008] Furthermore, the method for obtaining the authenticity coefficient includes: Taking the root of the blood vessel as the starting point, the level of the blood vessel area increases by 1 each time a branch point is passed, and each blood vessel area is graded; the corrosion indicators of all the blood vessel areas in the fundus blood vessel image are clustered; according to the difference characteristics of the corrosion indicators of all the blood vessel areas in each cluster, combined with the difference characteristics of the levels of all the blood vessel areas, a macular cluster is obtained; According to the difference between the corrosion index of each potential small blood vessel area and the corrosion index of the adjacent blood vessel areas on both sides, combined with the distance between the corrosion index of the potential small blood vessel area and the center of the macular cluster, and the small blood vessel coefficient of the potential small blood vessel area, the authenticity coefficient of each potential small blood vessel area is obtained; the difference between the corrosion index of the potential small blood vessel area and the corrosion index of the adjacent blood vessel areas on both sides is negatively correlated with the authenticity coefficient; the distance between the corrosion index of the potential small blood vessel area and the center of the macular cluster, and the small blood vessel coefficient, are both positively correlated with the authenticity coefficient.

[0009] Furthermore, the method for obtaining the macular cluster includes: In each of the clusters, the ratio of the variance of all the vascular area levels to the variance of the corrosion index is used as the macular possibility of each cluster; and the cluster with the largest macular possibility is selected as the macular cluster.

[0010] Furthermore, the method for obtaining the small blood vessel region includes: When the authenticity coefficient of the potential small blood vessel region is greater than a second preset threshold, the corresponding potential small blood vessel region is marked as a small blood vessel region.

[0011] Furthermore, the method for obtaining the enhancement coefficient includes: The ratio of the authenticity coefficient of the small blood vessel area to the corrosion index is normalized to be used as the enhancement coefficient corresponding to the small blood vessel area.

[0012] Furthermore, the method for performing image enhancement on the fundus vascular image according to the enhancement coefficient includes: The sum of the enhancement coefficient and a constant 1 is used as an enhancement parameter; The product of the contrast of the small blood vessel area and the enhancement parameter is used as the corrected contrast corresponding to the small blood vessel area; and the fundus blood vessel image is enhanced based on the corrected contrast of all the small blood vessel areas.

[0013] Furthermore, the method for acquiring the blood vessel region includes: Acquire a binary image of the fundus vascular image, extract a vascular skeleton of the binary image, and extract endpoints, branch points, and roots based on feature point detection; The blood vessel pixels between adjacent blood vessel branch points, between an end point and an adjacent blood vessel branch point, and between a root and an adjacent blood vessel branch point are recorded as one blood vessel region.

[0014] The present invention has the following beneficial effects: The present invention first obtains fundus vascular images and vascular regions to provide an analysis basis; further morphological corrosion is performed on each vascular region to facilitate optimization based on morphological corrosion, so as to more accurately locate small blood vessels and reduce macular interference; further corrosion indicators of each vascular region are obtained to quantify the changes in corroded pixel points, while reflecting the thickness of the blood vessels, providing a basis for locating small blood vessels; further, according to the difference characteristics of the corrosion indicators of the vascular regions in the preset neighborhood window of the target region, the small blood vessel coefficient of the target region is obtained, the potential small blood vessel region is obtained, the possibility of the target blood vessel being a small blood vessel is analyzed from a local perspective, and the vascular region is preliminarily screened for small blood vessels. Finally, the accurate small blood vessels are located, providing a basis for intelligent enhancement of the image; further, based on the small blood vessel coefficient, the possibility of potential small blood vessels belonging to real small blood vessels is evaluated from a global perspective, and the authenticity coefficient of the potential small blood vessel area is obtained to provide a basis for accurately locating the small blood vessel area; further, according to the authenticity coefficient, the small blood vessel area is obtained, and the small blood vessel area is accurately located to provide a basis for intelligent enhancement of the fundus blood vessel image; finally, according to the corrosion index and authenticity coefficient of each small blood vessel area, the enhancement coefficient of each small blood vessel area is obtained, and the fundus blood vessel image is enhanced to improve the details of the small blood vessel area in the fundus blood vessel image. The present invention utilizes the different characteristics of the changes in morphological corrosion of blood vessels of different diameters, and utilizes the similar characteristics of corrosion indicators between macular areas to screen out potential small blood vessel areas and evaluate the authenticity, and finally accurately locates the small blood vessel area and intelligently enhances the image to improve the image details. BRIEF DESCRIPTION OF THE DRAWINGS

[0015] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0016] Figure 1 A flowchart of a method for intelligently enhancing fundus vascular images provided by an embodiment of the present invention; Figure 2 A fundus blood vessel image provided by an embodiment of the present invention; Figure 3 An image of a fundus blood vessel image after morphological corrosion provided by an embodiment of the present invention; Figure 4 A schematic diagram of level division provided by an embodiment of the present invention. DETAILED DESCRIPTION

[0017] In order to further explain the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following is a detailed description of the specific implementation, structure, features and effects of an intelligent enhancement method for fundus vascular images proposed by the present invention in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0018] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this invention belongs.

[0019] The specific scheme of the intelligent enhancement method for fundus vascular images provided by the present invention is described in detail below with reference to the accompanying drawings.

[0020] See also Figure 1 , which shows a flow chart of a method for intelligent enhancement of fundus vascular images provided by an embodiment of the present invention, specifically comprising: Step S1: Acquire a fundus vascular image and obtain a vascular region; perform morphological corrosion on each vascular region.

[0021] The retina at the bottom of the human eyeball contains a large number of tiny blood vessels. The changes in the direction, diameter, width, shape and structure of these microvessels are all related to systemic diseases such as eye diseases, cardiovascular and cerebrovascular diseases, and hypertension. For example, diabetic retinopathy usually manifests as damage, occlusion or dilation of tiny blood vessels; hypertensive retinopathy may cause hardening or stenosis of retinal arterioles, and these changes often appear first in smaller blood vessels. The small blood vessels in the fundus vascular image are usually relatively small and complex in shape, and the blood flow rate is slow, which makes them difficult to detect or observe clearly in the fundus image, especially in the early stages of some diseases (such as diabetic retinopathy, glaucoma, hypertensive retinopathy, etc.). Small changes in small blood vessels may become the key to early diagnosis. Therefore, ensuring the image clarity and contrast of these small blood vessels is the key to improving diagnostic accuracy, and these small blood vessels need to be enhanced.

[0022] The embodiment of the present invention first obtains the fundus blood vessel image to provide the basis for analysis; please refer to Figure 2 , which shows a fundus vascular image provided by an embodiment of the present invention. To determine the tiny blood vessels in the fundus vascular image, firstly, the blood vessel region is acquired to determine the analysis range.

[0023] Preferably, in one embodiment of the present invention, the fundus vascular image is segmented by threshold value to obtain the vascular area; Figure 2Threshold segmentation is performed to obtain a binary image, and then the vascular skeleton is extracted using skeletonization technology. Based on feature point detection, endpoints and branch points are extracted, where an endpoint is a pixel point with only one adjacent pixel point, and a branch point is a pixel point with three or more adjacent pixels. The branch point with the largest diameter is taken as the root, and the endpoints, roots, and branch points are extracted. The vascular pixels between adjacent vascular branch points, between endpoints and adjacent vascular branch points, and between roots and adjacent vascular branch points are recorded as a vascular region.

[0024] It should be noted that the techniques for obtaining binary images by threshold segmentation, skeletonization technology, and endpoint detection have all been implemented. Specifically, the Otsu method or Frangi filter can be used to obtain the binary image, and the iterative shrinkage method or Voronoi skeleton can be used for skeletonization. In other embodiments of the present invention, the implementer can also train the convolutional neural network by manually annotating samples, and use the neural network to obtain the vascular area in the fundus vascular image, which will not be elaborated here.

[0025] Considering that the existing morphological corrosion may lead to problems such as excessive cutting, inaccurate judgment of breakpoints, and noise interference when extracting the vascular skeleton and locating small blood vessels, resulting in inaccurate judgment of small blood vessels and affecting image quality; in addition, the macular area may also interfere with the judgment of small blood vessels and affect the enhancement effect, so optimization based on morphological corrosion is performed to more accurately locate small blood vessels and reduce macular interference, so morphological corrosion is performed on each blood vessel area. Please refer to Figure 3 , which shows an image of a fundus blood vessel image after morphological corrosion provided by an embodiment of the present invention.

[0026] It should be noted that the morphological corrosion method is a technical means well known to those skilled in the art and will not be described in detail here.

[0027] Step S2: according to the change characteristics of the number of pixels before and after corrosion, the corrosion index of each blood vessel area is obtained; from the blood vessel root to any end point is recorded as a blood vessel branch; any blood vessel branch is selected as the target branch; any blood vessel area on the target branch is selected as the target area; according to the difference characteristics of the corrosion index of the blood vessel area within the preset neighborhood window of the target area, the small blood vessel coefficient of the target area is obtained; according to the small blood vessel coefficient, the potential small blood vessel area is obtained.

[0028] Considering that for blood vessels with larger or thicker diameters, the number of pixels in the blood vessel area changes significantly after morphological corrosion; and for blood vessel areas with thinner blood vessels, since the blood vessels themselves are thin, the number of pixels that can be removed by corrosion is limited, so the change in the number of pixels before and after corrosion is different between small blood vessels and thicker blood vessels. Therefore, according to the change characteristics of the number of pixels before and after corrosion, the corrosion index of each blood vessel area is obtained, and the change of corroded pixels is quantified, providing a basis for determining the small blood vessel area.

[0029] Preferably, in one embodiment of the present invention, considering that the greater the ratio of the change in the number of pixels before and after corrosion to the number of pixels before corrosion, the greater the impact of corrosion on the blood vessel area, the greater the corrosion index, and the more likely it is a thicker blood vessel; in each blood vessel area, the ratio of the change in the number of pixels before and after corrosion to the number of pixels before corrosion is used as the corrosion index of each blood vessel area.

[0030] It should be noted that before and after corrosion refer to before and after morphological corrosion, before corrosion refers to before morphological corrosion, and after corrosion refers to after morphological corrosion; the change in the number of pixels before and after corrosion refers to the absolute value of the difference between the number of pixels before corrosion and the number of pixels after corrosion; in another embodiment of the present invention, in each blood vessel area, the change in the number of pixels before and after corrosion is linearly normalized and used as the corrosion indicator of each blood vessel area.

[0031] In the fundus vascular skeleton, the diameter of fundus blood vessels changes regularly from thick to thin because the central artery starting from the optic disc branches and then enters different layers of the retina through gradual branching, eventually forming a tiny capillary network. However, in some special cases (such as blood vessel branching, pathological changes or physiological regulation, etc.), the retinal blood vessels may shrink in the short term, so in the process of the blood vessel diameter changing from thick to thin, there may be local abnormal thinning.

[0032] Since the corrosion indexes of blood vessel areas with different thicknesses are different, the possibility of small blood vessels can be analyzed by the changes in the corrosion indexes of the blood vessel areas on the blood vessel branches. From the root of the blood vessel to any end point, it is recorded as a blood vessel branch; any blood vessel branch is selected as the target branch; any blood vessel area on the target branch is selected as the target area; considering that when the blood vessel becomes abnormally thin locally, the corrosion index is quite different from the corrosion index of other blood vessel areas in the local neighborhood, and the smoothness of the change of the corrosion index in the local neighborhood is affected, the small blood vessel coefficient of the target area is obtained according to the difference characteristics of the corrosion index of the blood vessel area in the preset neighborhood window of the target area, and the possibility of the target blood vessel being a small blood vessel is analyzed from a local perspective, so as to provide a basis for the final accurate positioning of small blood vessels and intelligent enhancement of the image.

[0033] Preferably, in one embodiment of the present invention, considering that the greater the difference in corrosion index between the target area and the blood vessel areas on both sides within the preset neighborhood window, the worse the smoothness of the change in corrosion index within the local neighborhood, the less smooth the process of the blood vessel changing from thick to thin, indicating that the target area is more likely to be a small blood vessel area, and the smaller the small blood vessel coefficient is; at the same time, the greater the difference between the target area and the blood vessel areas on both sides relative to the maximum difference within the preset neighborhood window, indicating that the change in corrosion index of the target area is more prominent in the local neighborhood, and the small blood vessel characteristics are more obvious, based on this: According to the difference characteristics between the corrosion index of the target area and the corrosion index of the adjacent blood vessel areas on both sides, combined with the maximum difference in corrosion indexes among all blood vessel areas in the preset neighborhood window, the small blood vessel coefficient of the target area is obtained; the difference characteristics between the corrosion index of the target area and the corrosion index of the adjacent blood vessel areas on both sides are positively correlated with the small blood vessel coefficient; the maximum difference is negatively correlated with the small blood vessel coefficient.

[0034] As an example, the length of the preset neighborhood window is 5, that is, it contains 5 blood vessel regions, the target area is the center of the preset neighborhood window, and there are two adjacent blood vessel regions on both sides; within the preset neighborhood window of the target area, the blood vessel region on either side is regarded as the left adjacent blood vessel region, and the blood vessel region on the other side is regarded as the right adjacent blood vessel region; the calculation formula of the small blood vessel coefficient includes: ; Where i represents the serial number of the target area; represents the small blood vessel coefficient of the i-th target area; represents the corrosion index of the i-th target area; represents the mean value of the corrosion index of the left adjacent blood vessel area of ​​the i-th target area; represents the mean value of the corrosion index of the right adjacent blood vessel area of ​​the i-th target area; represents the maximum value of the corrosion index of all blood vessel regions within the preset neighborhood window of the i-th target region; It represents the minimum value of the corrosion index of all blood vessel regions within the preset neighborhood window of the i-th target region.

[0035] In the calculation formula of the small blood vessel coefficient, the corrosion index of the adjacent blood vessel area on one side is expressed by the average value; the difference characteristics between the corrosion index of the target blood vessel and the corrosion index of the adjacent area on one side are expressed by the absolute value of the difference, and then the average value of the absolute value of the difference corresponding to both sides is obtained to express the difference characteristics between the corrosion index of the target area and the corrosion index of the adjacent blood vessel areas on both sides; The larger the value is, the greater the difference in corrosion index between the target area and the blood vessel areas on both sides of the preset neighborhood window is, reflecting that the process of blood vessels changing from thick to thin is less smooth, indicating that the target area is more likely to be a small blood vessel area, and the smaller the blood vessel coefficient is, the greater the difference is. The larger it is, the greater the difference between the target area and the blood vessel areas on both sides relative to the maximum difference in the preset neighborhood window, which means that the change of corrosion index in the target area is more prominent in the local neighborhood, the characteristics of small blood vessels are more obvious, and the small blood vessel coefficient is larger.

[0036] It should be noted that when the target area has insufficient adjacent blood vessel areas, such as only one on one side, the corrosion index of the blood vessel area on this side is used to represent the corrosion index of the adjacent blood vessel area on the corresponding side; for example, when there is only one side, only the difference on one side is compared, and the average is no longer calculated, such as .

[0037] It should be noted that, considering that multiple vascular branches may pass through the same vascular region, in one embodiment of the present invention, the average value of the small vessel coefficients of all vascular branches corresponding to the vascular region is used as the final small vessel coefficient; in another embodiment of the present invention, the implementer may also select the longest vascular branch flowing through the same vascular region as the vascular branch belonging to this vascular region, and only calculate the small vessel coefficient in the longest vascular branch.

[0038] The small vessel coefficient measures the possibility that the target vessel is a small vessel from a local perspective. Therefore, the potential small vessel area is obtained according to the small vessel coefficient, the vessel area is preliminarily screened, and the analysis scope is narrowed.

[0039] Preferably, in one embodiment of the present invention, considering that the larger the small vessel coefficient is, the more likely the vessel region is a small vessel, when the small vessel coefficient is greater than or equal to a first preset threshold, the corresponding vessel region is marked as a potential small vessel region.

[0040] As an example, the first preset threshold is 0.5.

[0041] Step S3: According to the difference characteristics between the corrosion index of the potential small blood vessel area and the corrosion index of the adjacent blood vessel area, as well as the distribution characteristics of all corrosion indexes, combined with the small blood vessel coefficient, the authenticity coefficient of the potential small blood vessel area is obtained; according to the authenticity coefficient, the small blood vessel area is obtained; according to the corrosion index and authenticity coefficient of each small blood vessel area, the enhancement coefficient of each small blood vessel area is obtained.

[0042] Considering that the fundus vascular image contains the macular area, the macular area will also be corroded during morphological corrosion. The macular area may be identified as a breakpoint of small blood vessels or mistaken for a small blood vessel area. Therefore, further analysis of potential small blood vessel areas is required to eliminate interference from the macular area.

[0043] Taking into account that the corrosion indicators of different macular areas are relatively similar when they are corroded, while the corrosion indicators of blood vessel areas of different thicknesses in the blood vessel branches are quite different, this feature is used to further correct the small vessel coefficient in the blood vessel branches to which the potential small vessel area belongs according to the distribution characteristics of all corrosion indicators; considering that when small blood vessels appear in the blood vessel branches, the corrosion indicators of the small blood vessels and the adjacent blood vessel areas are quite different, the authenticity coefficient of the potential small blood vessel area is obtained based on the difference characteristics of the corrosion indicators of the potential small blood vessel area and the corrosion indicators of the adjacent blood vessel area, and the possibility of the potential small blood vessels belonging to the real small blood vessels is further evaluated from a global perspective, providing a basis for accurately locating the small blood vessel area.

[0044] Preferably, in one embodiment of the present invention, considering that the blood vessel gradually becomes thinner from the root to the end point, and the change is more obvious before and after the branch point, the root of the blood vessel is taken as the starting point, and the blood vessel area level increases by 1 every time a branch point is passed, and each blood vessel area is graded. Figure 4 , which shows a schematic diagram of level division provided by an embodiment of the present invention, Figure 4 Each line in the figure represents a blood vessel area, and the number represents the level of the blood vessel area. The level is used to indicate the distance between the blood vessel area and the root, and also indicates the thickness of the blood vessel. The higher the level, the farther from the root, and the thinner the blood vessel. Considering that the corrosion indexes of the macular area are relatively similar, that is, the corrosion indexes of macular areas of different levels are slightly different, the corrosion indexes of all vascular areas in the fundus vascular image are clustered; based on the difference characteristics of the corrosion indexes of all vascular areas in each cluster, combined with the difference characteristics of the levels of all vascular areas, the macular cluster is obtained to determine the cluster where the macular area is located; Considering that the actual small blood vessel area is far away from the macular cluster, and due to the abnormal thinning of the blood vessels, the corrosion index is likely to be lower than the corrosion index of the adjacent blood vessel areas on both sides. At the same time, the small blood vessel coefficient measures the possibility of the blood vessel area being a small blood vessel area from a local perspective. The larger the small blood vessel coefficient, the more likely the blood vessel area is a small blood vessel. Based on this: The authenticity coefficient of each potential small blood vessel area is obtained according to the difference between the corrosion index of each potential small blood vessel area and the corrosion index of the adjacent blood vessel areas on both sides, combined with the distance between the corrosion index of the potential small blood vessel area and the center of the macular cluster, and the small blood vessel coefficient of the potential small blood vessel area; the difference between the corrosion index of the potential small blood vessel area and the corrosion index of the adjacent blood vessel areas on both sides is negatively correlated with the authenticity coefficient; the distance between the corrosion index of the potential small blood vessel area and the center of the macular cluster, as well as the small blood vessel coefficient, are both positively correlated with the authenticity coefficient.

[0045] The calculation formula of the authenticity coefficient includes: ; Among them, k represents the serial number of potential small blood vessels; represents the authenticity coefficient of the kth potential small blood vessel; represents the linear normalization function; represents the small vessel coefficient of the kth potential small vessel; represents the distance between the corrosion index of the kth potential small blood vessel and the center of the macula cluster; It represents the exponential function with the natural constant e as the base; represents the corrosion index of the kth potential small blood vessel; It represents the corrosion index of the vessel area with the smallest level adjacent to the kth potential small vessel; Represents the corrosion index of the blood vessel area with the largest level adjacent to the kth potential small blood vessel.

[0046] In the calculation formula of the authenticity coefficient, the negative correlation mapping function The corrosion index difference is negatively correlated and mapped, and x represents the independent variable; middle, and The smaller it is, the smaller the corrosion index of the potential small blood vessel area is relative to the corrosion index of the adjacent blood vessel areas on both sides, the more obvious the abnormal thinning characteristics of the potential small blood vessel area are, the more likely it is a real small blood vessel, and the larger the true coefficient is; The larger it is, the greater the difference between the potential small blood vessel area and the macular clustering in the macular area, the less likely it is the macular area, and the greater the authenticity coefficient; The larger it is, the more likely it is that the vascular area is a small blood vessel from a local perspective, and the greater the authenticity coefficient.

[0047] It should be noted that, in one embodiment of the present invention, the k-means clustering algorithm is used for clustering, the cluster center is the mean of all corrosion indicators in the cluster, and the distance from the corrosion indicator to the cluster center is the absolute value of the difference between the corrosion indicator and the cluster center; when there is only one adjacent blood vessel area for potential small blood vessels, only the difference corresponding to one adjacent blood vessel is calculated, and the average is not taken. hour, .

[0048] In other embodiments of the present invention, implementers may also use other clustering algorithms such as DBSCAN clustering, which and the k-means clustering algorithm are both existing technologies and will not be described in detail.

[0049] Preferably, in one embodiment of the present invention, considering that the more chaotic the level of the vascular area within the cluster is and the more concentrated the corrosion index is, the more the cluster conforms to the characteristics of the macular area of ​​different levels and similar corrosion indicators; and considering that the larger the variance is, the larger the data fluctuation is and the more chaotic it is, so in each cluster, the ratio of the variance of all vascular area levels to the variance of the corrosion index is used as the macular possibility of each cluster; the cluster with the largest macular possibility is selected as the macular cluster.

[0050] It should be noted that, usually, the corrosion indicators of different vascular areas have certain differences, that is, the variance of the corrosion indicators is not zero. The implementer can also use a preset positive parameter for zero division, such as 0.01, and use the variance of all vascular area levels as the numerator, the sum of the variance of the corrosion indicator and 0.01 as the denominator, and use the fractional ratio as the macular possibility of each cluster.

[0051] It should be noted that when the potential small blood vessel region is close to the branch point, there may be multiple adjacent blood vessel regions on one side, e.g. Figure 4 For example, there is a level 2 vascular region in the middle, and there are three adjacent level 3 vascular regions on one side. At this time, the implementer can stipulate to calculate the average value of the corrosion indexes of the three adjacent vascular regions as the corrosion index of the vascular region with the largest level adjacent to potential small blood vessels; or the implementer can stipulate to select the vascular region on the longest vascular branch, that is, select the corrosion index of the rightmost vascular region as the corrosion index of the vascular region with the largest level adjacent to potential small blood vessels.

[0052] On the basis of the locally evaluated small vessel coefficient, the authenticity of the potential small vessel area is further evaluated from a global perspective to obtain the authenticity coefficient. Then, based on the authenticity coefficient, the small vessel area is obtained and accurately located, providing a basis for intelligent enhancement of fundus vascular images.

[0053] Preferably, in one embodiment of the present invention, considering that the larger the authenticity coefficient is, the more likely the potential small blood vessel region is to be a real small blood vessel region, so when the authenticity coefficient of the potential small blood vessel region is greater than the second preset threshold, the corresponding potential small blood vessel region is marked as a small blood vessel region.

[0054] As an example, the second preset threshold is 0.5.

[0055] The corrosion index reflects the change in the number of pixels of small blood vessels before and after corrosion, reflecting the thickness of the small blood vessels; the authenticity coefficient characterizes the authenticity of the small blood vessels. Therefore, according to the corrosion index and authenticity coefficient of each small blood vessel area, the enhancement coefficient of each small blood vessel area is obtained, providing an enhancement basis for accurately enhancing small blood vessels.

[0056] Preferably, in one embodiment of the present invention, considering that the larger the authenticity coefficient is, the more realistic the small blood vessels are; the smaller the corrosion index is, the thinner the small blood vessels are, and the thinner and more realistic the small blood vessels are, the more they need to be enhanced, so the ratio of the authenticity coefficient of the small blood vessel area to the corrosion index is normalized and used as the enhancement coefficient of the corresponding small blood vessel area.

[0057] It should be noted that a linear normalization method is specifically adopted; in other embodiments of the present invention, the implementer may also perform negative correlation mapping on the corrosion index through a negative correlation mapping function such as exp(-x), linearly normalize the product of the negative correlation mapping value and the authenticity coefficient, and use the normalized result as the enhancement coefficient of the corresponding small blood vessel area.

[0058] Step S4: performing image enhancement on the fundus vascular image according to the enhancement coefficient.

[0059] The enhancement coefficient characterizes the degree of enhancement of small blood vessels. After determining the enhancement coefficients of all small blood vessels, the fundus vascular image can be enhanced according to the enhancement coefficient to improve the details of the small blood vessel area in the fundus vascular image, thereby providing more accurate imaging data for clinical diagnosis and prevention of related diseases.

[0060] Preferably, in one embodiment of the present invention, the sum of the enhancement coefficient and the constant 1 is used as the enhancement parameter; The product of the contrast of the small blood vessel area and the enhancement parameter is used as the corrected contrast of the corresponding small blood vessel area; based on the corrected contrast of all small blood vessel areas, the fundus blood vessel image is enhanced.

[0061] In other embodiments of the present invention, the implementer may also map the enhancement coefficient to other positive intervals, such as mapping to [0, 2]. In this case, the enhancement parameter is [1, 3], and the maximum corrected contrast is 3 times the original contrast.

[0062] It should be noted that obtaining contrast and contrast enhancement methods are technical means well known to those skilled in the art. In one embodiment of the present invention, a local comparison algorithm can be used to achieve local contrast enhancement; in other embodiments of the present invention, the implementer can obtain the average grayscale value of the small blood vessel area as the foreground grayscale value, use other areas in the fundus vascular image as the background area, obtain the average grayscale value of the background area as the background grayscale value, and use the difference between the foreground grayscale value and the background grayscale value and the ratio of the sum of the foreground grayscale value and the background grayscale value as the contrast; after calculating the corrected contrast, the pixel values ​​of the pixels in the small blood vessel area are linearly transformed, which will not be repeated here.

[0063] In other embodiments of the present invention, a generative adversarial network can also be used to perform texture enhancement on small blood vessel areas. When training the model, the enhancement coefficient is used as a guiding parameter, and attention is paid to the details of the small blood vessel areas. The generative adversarial network is already an existing technology and will not be described in detail.

[0064] In summary, in view of the technical problems of inaccurate positioning of small blood vessel areas and unsatisfactory image enhancement effects, the present invention proposes an intelligent enhancement method for fundus blood vessel images. The present invention first obtains fundus blood vessel images and blood vessel areas; further obtains corrosion indicators of each blood vessel area according to the change characteristics of the number of pixels before and after corrosion; further obtains the small blood vessel coefficient of the target area according to the difference characteristics of the corrosion indicators of the blood vessel areas in the preset neighborhood window of the target area, and obtains the potential small blood vessel area; further obtains the authenticity coefficient of the potential small blood vessel area according to the difference characteristics of the corrosion indicators of the potential small blood vessel area and the corrosion indicators of the adjacent blood vessel areas, as well as the distribution characteristics of all corrosion indicators, combined with the small blood vessel coefficient, and obtains the small blood vessel area; finally, according to the corrosion indicators and authenticity coefficients of each small blood vessel area, the enhancement coefficient of each small blood vessel area is obtained, and the fundus blood vessel image is enhanced. The present invention utilizes the different characteristics of the changes of blood vessels of different diameters after morphological corrosion, and utilizes the similar characteristics of corrosion indicators between macular areas to screen out potential small blood vessel areas and evaluate the authenticity, and finally accurately locates the small blood vessel area and intelligently enhances the image to improve the image details.

[0065] It should be noted that the sequence of the above embodiments of the present invention is only for description and does not represent the advantages and disadvantages of the embodiments. The processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0066] The various embodiments in this specification are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

Claims

1. A method for intelligent enhancement of fundus vascular images, characterized in that: The method comprises: Acquire fundus vascular images and vascular regions; perform morphological corrosion on each of the vascular regions; According to the change characteristics of the number of pixel points before and after corrosion, the corrosion index of each blood vessel area is obtained; from the blood vessel root to any end point is recorded as a blood vessel branch; any blood vessel branch is selected as a target branch; any blood vessel area on the target branch is selected as a target area; according to the difference characteristics of the corrosion index of the blood vessel area in the preset neighborhood window of the target area, the small blood vessel coefficient of the target area is obtained; according to the small blood vessel coefficient, the potential small blood vessel area is obtained; According to the difference characteristics between the corrosion index of the potential small blood vessel area and the corrosion index of the adjacent blood vessel area, and the distribution characteristics of all the corrosion indexes, combined with the small blood vessel coefficient, the authenticity coefficient of the potential small blood vessel area is obtained; according to the authenticity coefficient, the small blood vessel area is obtained; according to the corrosion index and the authenticity coefficient of each small blood vessel area, the enhancement coefficient of each small blood vessel area is obtained; The fundus blood vessel image is enhanced according to the enhancement coefficient.

2. The method for intelligent enhancement of fundus vascular images according to claim 1, characterized in that: The method for obtaining the corrosion index includes: In each of the blood vessel regions, the ratio of the change in the number of pixel points before and after corrosion to the number of pixel points before corrosion is used as a corrosion index for each of the blood vessel regions.

3. The method for intelligent enhancement of fundus vascular images according to claim 1, characterized in that: The method for obtaining the small blood vessel coefficient includes: The preset neighborhood window is centered on the target area; According to the difference characteristics between the corrosion index of the target area and the corrosion index of the adjacent blood vessel areas on both sides, combined with the maximum difference of the corrosion index between all the blood vessel areas in the preset neighborhood window, the small blood vessel coefficient of the target area is obtained; the difference characteristics between the corrosion index of the target area and the corrosion index of the adjacent blood vessel areas on both sides are positively correlated with the small blood vessel coefficient; the maximum difference is negatively correlated with the small blood vessel coefficient.

4. The method for intelligent enhancement of fundus vascular images according to claim 3, characterized in that: The method for obtaining the potential small blood vessel area comprises: When the small blood vessel coefficient is greater than or equal to a first preset threshold, the corresponding blood vessel region is marked as a potential small blood vessel region.

5. The method for intelligent enhancement of fundus vascular images according to claim 1, characterized in that: The method for obtaining the authenticity coefficient includes: Taking the root of the blood vessel as the starting point, the level of the blood vessel area increases by 1 each time a branch point is passed, and each blood vessel area is graded; the corrosion indicators of all the blood vessel areas in the fundus blood vessel image are clustered; according to the difference characteristics of the corrosion indicators of all the blood vessel areas in each cluster, combined with the difference characteristics of the levels of all the blood vessel areas, a macular cluster is obtained; According to the difference between the corrosion index of each potential small blood vessel area and the corrosion index of the adjacent blood vessel areas on both sides, combined with the distance between the corrosion index of the potential small blood vessel area and the center of the macular cluster, and the small blood vessel coefficient of the potential small blood vessel area, the authenticity coefficient of each potential small blood vessel area is obtained; the difference between the corrosion index of the potential small blood vessel area and the corrosion index of the adjacent blood vessel areas on both sides is negatively correlated with the authenticity coefficient; the distance between the corrosion index of the potential small blood vessel area and the center of the macular cluster, and the small blood vessel coefficient, are both positively correlated with the authenticity coefficient.

6. The method for intelligent enhancement of fundus vascular images according to claim 5, characterized in that: The method for obtaining the macular cluster includes: In each of the clusters, the ratio of the variance of all the vascular area levels to the variance of the corrosion index is used as the macular possibility of each cluster; and the cluster with the largest macular possibility is selected as the macular cluster.

7. The method for intelligent enhancement of fundus vascular images according to claim 5, characterized in that: The method for obtaining the small blood vessel region comprises: When the authenticity coefficient of the potential small blood vessel region is greater than a second preset threshold, the corresponding potential small blood vessel region is marked as a small blood vessel region.

8. The method for intelligent enhancement of fundus vascular images according to claim 1, characterized in that: The method for obtaining the enhancement coefficient includes: The ratio of the authenticity coefficient of the small blood vessel area to the corrosion index is normalized to be used as the enhancement coefficient corresponding to the small blood vessel area.

9. The method for intelligent enhancement of fundus vascular images according to claim 8, characterized in that: The method for performing image enhancement on the fundus vascular image according to the enhancement coefficient comprises: The sum of the enhancement coefficient and a constant 1 is used as an enhancement parameter; The product of the contrast of the small blood vessel area and the enhancement parameter is used as the corrected contrast corresponding to the small blood vessel area; and the fundus blood vessel image is enhanced based on the corrected contrast of all the small blood vessel areas.

10. The method for intelligent enhancement of fundus vascular images according to claim 1, characterized in that: The method for acquiring the blood vessel region comprises: Acquire a binary image of the fundus vascular image, extract a vascular skeleton of the binary image, and extract endpoints, branch points, and roots based on feature point detection; The blood vessel pixels between adjacent blood vessel branch points, between an end point and an adjacent blood vessel branch point, and between a root and an adjacent blood vessel branch point are recorded as one blood vessel region.

Images