Method, device, computer storage medium and electronic device for processing eye vein feature point information

Abstract

The present invention discloses a method, device, computer storage medium and electronic device for processing eye vein feature point information. The eye vein feature point information processing method includes: obtaining an input eye image, extracting the blood vessel area of ​​the eye image; extracting feature points in the eye image; establishing a vein feature data unit based on the degree of overlap between a line segment formed between at least two feature points and the blood vessel area of ​​the eye image; calculating the vector of each line segment formed by the ordered reference feature points and key feature points in each vein feature data unit, and adding the vector to the vein feature data unit; based on the vector formed by each two key feature points in the vein feature data unit, using an expansion algorithm to extract the blood vessel area corresponding to each vein feature data unit. The eye blood vessel feature point information processing method of the present invention can systematically and orderly screen and sort feature points, and establish a corresponding relationship with the blood vessel image. The method is simple and fast, and can accurately establish machine recognition of blood vessel image information through feature points, thereby realizing the effective application of automatic eye image processing equipment.

Description

Technical Field

[0001] The present invention relates to the field of human eye image recognition, and in particular to a method and device for processing eye vein feature point information, a computer storage medium, and electronic equipment. Background Art

[0002] The 14 meridians and the eight extraordinary meridians of the human body are all directly or indirectly related to the eyes. Therefore, if there are abnormal changes in the functions of the internal organs, they will be manifested in the corresponding parts of the eyes, and different signs will appear, such as spots and astringency in the eyes. Eye information recognition can determine the changes in the corresponding internal organs, and then assist in regulating the functions of the corresponding internal organs. Therefore, identifying the health status of the human body through eye diagnosis, and adjusting, intervening and treating in advance are considered to be very reliable, practical and effective methods for human health.

[0003] Currently, most eye diagnostic equipment is completed through human observation and experience-based judgment, which has a great limiting effect on efficiency, the number of patients treated, and the promotion and influence expansion of Chinese medicine eye diagnosis. The only eye diagnosis equipment is unable to accurately determine the required results, the image processing algorithm is fuzzy, the modeling is inaccurate, and the degree of automation is poor, which has limited its practical application. Therefore, the current application of eye information collection and processing in the medical field faces common problems that need to be solved urgently, such as reliable eye diagnosis equipment, accurate acquisition of the information required for eye diagnosis, and the ability to achieve automation and high efficiency. Summary of the Invention

[0004] In view of the above shortcomings of the prior art, the present invention provides a method, device, computer equipment and storage medium for processing ocular blood vessel feature point information, which can improve the above shortcomings.

[0005] As one aspect of the present invention, the present invention provides a method for processing eye vein feature point information, which is used to process eye vein image information, wherein the method includes:

[0006] Acquire an input eye image and extract a blood vessel region of the eye image;

[0007] Extracting feature points from the eye image;

[0008] establishing a vein feature data unit based on a degree of overlap between a line segment formed between at least two feature points and a blood vessel region of the eye image;

[0009] Calculating the vector of each line segment formed by the ordered reference feature points and key feature points in each venation feature data unit, and adding the vector to the venation feature data unit, wherein the reference feature points are feature points selected and determined based on the distance from the edge of the blood vessel region; and the key feature points are feature points selected and determined based on the degree of overlap between the line segment formed by the at least one reference feature point and the blood vessel region;

[0010] Based on the vector formed by connecting every two key feature points in the venous feature data unit, a dilation algorithm is used to calculate the blood vessel area corresponding to each venous feature data unit.

[0011] In one embodiment, the extracting the blood vessel region of the eye image specifically includes extracting the blood vessel region in the eye image using an adaptive binarization method.

[0012] In one embodiment, the extracting feature points from the eye image specifically includes extracting feature points from the blood vessel image using an ORB feature algorithm or a deep learning SuperPonit model.

[0013] In one embodiment, the venation feature data unit is established based on the degree of overlap between a line segment formed between at least two feature points and the blood vessel area of ​​the eye image, specifically comprising: selecting a reference feature point from a set of feature points extracted from the eye image based on the distance between the feature point and the edge of the blood vessel area, and establishing the venation feature data unit based on the degree of overlap between a line segment formed between feature points other than the reference feature point and the reference feature point and the blood vessel area.

[0014] In one embodiment, the venation feature data unit is established by selecting a reference feature point from the feature point set extracted from the eye image based on the distance between the feature point and the edge of the blood vessel region, and by establishing a venation feature data unit based on the degree of overlap between a line segment formed by other feature points other than the reference feature point and the reference feature point and the blood vessel region, specifically including:

[0015] Step a: extracting feature points in the eye image whose distance to the edge of the blood vessel region is less than a first preset threshold, and forming a reference feature point set with all M feature points whose distance to the edge of the blood vessel region is less than the first preset threshold, where M is a positive integer greater than or equal to 1;

[0016] Step b: selecting a feature point with the shortest distance from the edge of the blood vessel region in the reference feature point set as a starting feature point, and using the starting feature point as a current marking point;

[0017] Step c: determining a feature point adjacent to the starting feature point outside the reference feature point set in the blood vessel region as a second key feature point;

[0018] Step d, determining the degree of overlap between the line segment formed by the starting feature point and the second key feature point and the blood vessel region; if the degree of overlap is higher than a second threshold, adding the second key feature point to the venation feature data unit; if the degree of overlap is lower than the second threshold, excluding the second key feature point;

[0019] Step e, repeating steps cd until all feature points outside the reference feature point set of the vascular image are traversed based on the starting feature point, and the starting feature point and all second key feature points satisfying a degree of overlap higher than a second threshold are added to the venation feature data unit;

[0020] Step f, repeating steps b to e until the traversal of the vein feature data units based on all M reference feature points is completed.

[0021] In one embodiment, the first preset threshold is the maximum distance between the 5% of feature points closest to the edge of the blood vessel region among all feature points and the edge of the blood vessel region.

[0022] In one embodiment, the calculating of the vector of each line segment formed by the ordered reference feature points and key feature points in each context feature data unit and adding the vector to the context feature data unit further includes: determining the starting feature point from the reference feature points, and the line segment formed by the key feature points in step d corresponding to the starting feature point and having a degree of overlap higher than a second threshold, generating a vector, and adding a plurality of the vectors to the context feature data unit.

[0023] The present invention also provides an eye vein feature point information processing device for processing eye vein image information, wherein the device comprises:

[0024] a first extraction unit, configured to acquire an input eye image and extract a blood vessel region of the eye image;

[0025] a second extraction unit, configured to extract feature points from the eye image;

[0026] a processing unit, which establishes a vein feature data unit based on a degree of overlap between a line segment formed between at least two feature points and a blood vessel region of the eye image;

[0027] a first calculation unit, calculating a vector of each line segment formed by connecting the ordered reference feature points and key feature points in each venation feature data unit, and adding the vector to the venation feature data unit, wherein the reference feature points are feature points selected and determined based on the distance from the edge of the blood vessel region; and the key feature points are feature points selected and determined based on the degree of overlap between the line segment connected to the at least one reference feature point and the blood vessel region;

[0028] The second calculation unit is configured to calculate the blood vessel area corresponding to each venous feature data unit by using a dilation algorithm based on a vector formed by connecting every two key feature points in the venous feature data unit.

[0029] The present invention also provides a computer storage medium storing a plurality of instructions, wherein the instructions are suitable for being loaded by a processor and executed by the steps of the eye vein feature point information processing method as in any of the above-mentioned embodiments.

[0030] The present invention also provides an electronic device comprising: a processor and a memory; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the steps of the eye vein feature point information processing method as in any of the above-mentioned embodiments.

[0031] The present invention sets a reference feature point set based on a venous image from which vascular regions and feature points are extracted, selects a starting feature point through a layer-by-layer approximation algorithm, compares the nearest second key feature point outside the reference feature point set with the degree of overlap between the line segment formed between the starting feature point and the vascular region, adds feature points that meet the overlap judgment condition, establishes a venous feature data unit, and adds the line segment vector formed by connecting ordered key feature points in the venous feature point data unit to the venous feature data unit. In this way, the venous feature data unit can automatically, efficiently and accurately screen and sort scattered vascular region feature points based on the correlation of the blood vessels, and correspond the starting feature point and the second key feature point in the venous feature data unit that meet the overlap judgment standard to the actual blood vessels in the vascular image through an expansion algorithm, thereby achieving good equipment automation and precise detection effects in the eye diagnosis process or by integrating into eye diagnosis equipment, thereby realizing the expanded application of eye diagnosis equipment.

[0032] Other features and advantages of the present invention will be described in the following description, and in part will become apparent from the description, or will be understood by practicing the present invention. The purposes and other advantages of the present invention can be realized and obtained by the structures particularly pointed out in the written description, claims, and drawings. BRIEF DESCRIPTION OF THE DRAWINGS

[0033] The drawings described herein are used to provide a further understanding of the present invention and constitute a part of the present invention. The exemplary embodiments of the present invention and their descriptions are used to explain the present invention and do not constitute an improper limitation of the present invention. In the drawings:

[0034] Figure 1 Flowchart of a method for processing eye vein feature point information according to embodiment 1 of the present invention

[0035] Figure 2 Flowchart for establishing the context feature data unit in embodiment 1 of the present invention

[0036] Figure 3 Schematic diagram of the eye blood vessel feature point information processing device in the present invention

[0037] Figure 4 Schematic diagram of the electronic device structure in the present invention DETAILED DESCRIPTION

[0038] The preferred embodiments of the present invention are described below in conjunction with the drawings. It should be understood that the preferred embodiments described herein are only used to illustrate and explain the present invention and are not used to limit the present invention. In addition, the embodiments of the present invention and the features in the embodiments may be combined with each other if there is no conflict.

[0039] In order to facilitate the understanding of the embodiments of the present invention, a brief description is first given of the eye information processing equipment in the related art. With respect to the field of human eye image recognition, especially in the automatic eye information processing equipment, the demand for accurate processing of the core identification element, namely the eye vein features including blood vessels, is still not met by the existing technology. The eye diagnosis equipment in the existing technology pays more attention to technical improvements in the hardware composition and arrangement of the equipment, including cameras, light sources, etc. After obtaining the photo through the camera mechanism, the equipment semi-automatically processes part of the information only through rough identification of the white part and the eyeball, and then feeds back the semi-automatically processed picture and the result. This is still far from the current requirements of automation, intelligence, convenience and accuracy of the equipment, and cannot help the rapid popularization of the equipment, which affects the application of the current eye information processing technology.

[0040] There are also related existing technologies that use machine learning to identify eye information and judge symptoms, but most of them directly input eye images collected by equipment with different parameters and uneven quality after selecting a large machine model, and let the machine learn. This technical route has poor training set quality and it is difficult to form an ultra-large-scale training set. The model selection is difficult, there is no idea of ​​pre-processing the original image, and the training operation is too complicated. Therefore, it is difficult to produce usable results in actual applications.

[0041] To solve this problem, embodiment 1 of the present invention discloses a method for processing eye vein feature point information for processing eye image information, specifically comprising:

[0042] Step 101: Obtain an input eye image and extract a blood vessel area of ​​the eye image.

[0043] As the basic data for eye diagnosis equipment identification and calculation, step 101 obtains an eye image of the person to be examined through a camera assembly and calculates the vascular area in the eye image. In the implementation of the present invention, the human eye image can be used as an auxiliary support for fixing and presetting the human eye state of different people through an auxiliary device. A camera is used to photograph the target area of ​​the target human eye to obtain the human eye image. In order to support subsequent image processing and algorithm model recognition, in the embodiment of the present invention, a camera that can clearly capture the details and color of the blood vessels around the eye, including the eye itself, is preferred. A camera with corresponding high pixel and resolution, such as a SLR camera, a smartphone camera, an industrial camera, etc., can be selected.

[0044] In embodiments of the present invention, extracting vascular regions from eye images specifically involves employing an adaptive binarization method to extract vascular regions within the eye image. The white and clear regions of the eye image contain a greater number of blood vessels, which are scattered and irregular in shape. However, due to the significant difference in image quality compared to the white and clear regions, the present invention employs an algorithm to perform special processing on the vascular regions within the eye image. This precise extraction of vascular region information allows for processing of vascular feature point information, ensuring accurate acquisition of vascular feature information.

[0045] In the embodiment of the present invention, the adaptive binarization is used to extract the vascular information of the eye image, which is to convert the grayscale value of the image into only 0 and 255, that is, black or white, with black (0) as the background (dark area) and white (255) as the foreground, i.e., the target area (bright area). Generally, a threshold is set, and all pixels greater than the threshold are set to 0 or 255, and all pixels less than the threshold are set to a grayscale, thus completing the binarization of the image. Common methods can be divided into two categories: fixed threshold method and adaptive threshold method. Fixed threshold method: means setting a threshold according to the threshold type and then performing binarization. Adaptive threshold method: means setting a threshold according to the average or weighted pixels of the neighborhood module of each pixel point and then performing binarization. Therefore, for the case where the image information of the eye diagnosis device is significantly different from the vascular information compared to the white eye, the use of the binarization method for extraction can achieve accurate and convenient image extraction effects.

[0046] Step 102: Extract feature points from the eye image.

[0047] To more accurately reconstruct, restore, and identify vascular information in the white part of the eye image via a computer, embodiments of the present invention employ a method for expressing vascular information using feature points. Extracting feature points from eye images lays the foundation for subsequent feature point expression, reconstruction, restoration, and identification of vascular information. As a preferred embodiment, embodiments of the present invention employ either the ORB feature algorithm or the deep learning SuperPoint model to extract feature points from eye images. The ORB feature algorithm can be used to quickly create feature vectors from key points in an image, which can be used to identify objects within the image. Fast and Brief algorithms can be used as feature detection and vector creation algorithms, respectively. The ORB feature algorithm offers the advantages of speed and low resource consumption, but suffers from limitations in accuracy. The deep learning SuperPoint algorithm, when trained on a sufficient training set, can achieve higher feature point recognition accuracy. In actual application, skilled artisans can choose between the two algorithms based on their specific needs, considering speed, simplicity, and accuracy.

[0048] Step 103: establishing a vein feature data unit based on the degree of overlap between a line segment connecting at least two feature points and a blood vessel region of the eye image.

[0049] It should be noted that in Example 1 of the present invention, an algorithm is used to identify and determine the feature points obtained in step 102, and target feature points are identified and selected by matching the degree of overlap with the blood vessel region of the eye image. This provides a target feature point set for subsequent calculations and is an important implementation method for accurately identifying and reconstructing blood vessels by a computer. In this embodiment of the present invention, the establishment of a venous feature data unit based on the degree of overlap between a line segment formed between at least two feature points and the blood vessel region of the eye image specifically includes: selecting a reference feature point from the set of feature points extracted from the eye image based on the distance between the feature point and the edge of the blood vessel region, and establishing the venous feature data unit based on the degree of overlap between a line segment formed between feature points other than the reference feature point and the blood vessel region.

[0050] In Example 1 of the present invention, by setting the degree of overlap between a line segment formed by connecting any two feature points and the blood vessel image area, target feature points are identified and screened, and a venous feature data unit is established based on the set of target feature points. By extracting information of feature points that can be recognized and processed by a computer, blood vessel information in the image is represented. On the basis of ensuring rapid computer recognition, accurate matching with the blood vessels can be achieved, so that the computer can process according to the theory of traditional Chinese medicine eye diagnosis and identify eye information.

[0051] Step 104: Calculate the vector of each line segment formed by the ordered reference feature points and key feature points in each context feature data unit, and add it to the context feature data unit.

[0052] In Example 1 of the present invention, in addition to the reference feature points, the venation feature data unit also includes the vector of each line segment connected by the ordered reference feature points and the key feature points in each venation feature data unit. The vector of each line segment connected by the ordered reference feature points and the key feature points is added to the venation feature data unit, and the computer subsequently processes and identifies the required vascular area according to the theory of traditional Chinese medicine eye diagnosis to restore the vector, thereby realizing the subsequent automatic information extraction of the vascular area.

[0053] It should be noted that, in the embodiments of the present invention, the Traditional Chinese Medicine eye information processing theory is a method for zoning the white of the human eye's internal organs, specifically a theory of eye acupuncture from Professor Peng Jingshan, a renowned senior TCM practitioner. This theory is part of the National Key Basic Research and Development Program of Traditional Chinese Medicine (973 Program), with project number 2007CB512707. It automatically identifies eye information by dividing the left and right eyes into eight zones and thirteen acupoints, based on the theory of eye zoning, and the correspondence between vein characteristics and internal organs.

[0054] In the embodiment of the present invention, the reference feature points are feature points selected and determined based on the distance from the edge of the blood vessel region; the key feature points are feature points selected and determined based on the degree of overlap between the line segment connected to the at least one reference feature point and the blood vessel region.

[0055] Step 105: Based on the vector formed by connecting every two key feature points in the venation feature data unit, a dilation algorithm is used to calculate the blood vessel area corresponding to each venation feature data unit.

[0056] In Example 1 of the present invention, step 104 adds the reference feature points and each line segment vector formed by connecting the ordered reference feature points and the key feature points to the venation feature data unit, thereby enabling the blood vessel area to be filled based on the line segment vectors formed by connecting the reference feature points and the key feature points, enabling the computer to automatically and accurately obtain blood vessel image information, and automatically and accurately process and identify the eye image.

[0057] It should be noted that, in an embodiment of the present invention, the key feature points are feature points selected and determined based on the degree of overlap between the line segments connected with the at least one reference feature point and the vascular area. Each line segment connected with the ordered reference feature points and the key feature points can be the order of the line segments connected between at least one reference feature that meets the requirements of overlap with the vascular area and the key feature points.

[0058] The human eye vein characteristics described in the embodiment of the present invention include at least the blood vessel image information of color, thickness, and spots. In addition to the above information, other blood vessel and nerve image information that can reflect the human eye veins can also be included in the present invention.

[0059] In the first embodiment of the present invention, based on a human eye image from which blood vessel regions and feature points are extracted, a reference feature point set is set using a layer-by-layer approximation algorithm, a starting feature point is selected, and the nearest second key feature point outside the reference feature point set is compared one by one with the degree of overlap between a line segment formed between the starting feature point and the blood vessel region. Feature points that meet the overlap judgment criteria are added to establish a venous feature data unit, and line segment vectors formed by connecting ordered key feature points in the venous feature point data unit are added to the venous feature data unit. In this way, the venous feature data unit can automatically, efficiently, and accurately screen and sort scattered vascular region feature points based on their relevance to the blood vessels, and map the starting feature points and second key feature points in the venous feature data unit that meet the overlap judgment criteria to actual blood vessels in the blood vessel image using an expansion algorithm. This allows for excellent equipment automation and precise detection in an eye information processing process or when integrated into an eye information analysis device, thereby achieving expanded application of the device.

[0060] Attachment Figure 2 This is a flow chart of an implementation method for establishing the vascular feature data unit described in Example 1 of the present invention. Specifically, it includes:

[0061] Step a: extract feature points in the eye image whose distance to the edge of the blood vessel area is less than a first preset threshold, and form a reference feature point set with all M feature points whose distance to the edge of the blood vessel area is less than the first preset threshold, where M is a positive integer greater than or equal to 1.

[0062] It should be noted that the establishment of the ventricle feature data unit first requires calculation and determination of the reference feature points. In the embodiment of the present invention, the reference feature points are selected from all feature points by the relationship between the distance to the edge of the vascular area and the first preset threshold, wherein the first preset threshold can be selected and set by technicians in this field according to the actual conditions of different human bodies and eyes. In one example, the first preset threshold is the maximum distance between the 5% of feature points closest to the edge of the vascular area among all feature points and the edge of the vascular area, or it can be the distance obtained after weighting the farthest distance and the closest distance to the vascular area among all feature points as the first preset threshold. In the embodiment of the present invention, there is no limitation on this, as long as a certain number of reference feature points that are closer to the edge of the vascular area than all feature points can be selected, key feature points that meet the calculation requirements can be calculated with the reference feature points as the starting point, and the ventricle feature data unit can be obtained.

[0063] Step b: Selecting a feature point with the shortest distance from the edge of the blood vessel region in the reference feature point set as a starting feature point, and using the starting feature point as a current marking point.

[0064] Since vascular information needs to be represented in the form of a vector in order to be automatically calculated and recognized by a computer, it is necessary to select at least two target feature points during the feature point calculation and expression process, and use the line segment formed by them to express the blood vessel. In the embodiment of the present invention, after obtaining a set of reference feature points whose distance from the edge of the vascular region is less than a first preset threshold in step a, how to traverse and form feature points that can form a line segment and accurately represent the vascular information is solved by the implementation method of establishing the venous feature data unit described in the embodiment of the present invention. In step b, the feature point with the smallest distance from the edge of the vascular region in the reference feature point set is selected as the starting feature point, and the starting feature point is used as the current marking point, which provides a basic reference for subsequent traversal and is also a relatively simple and convenient calculation method.

[0065] Step c: Determine a feature point that is closest to the starting feature point outside the reference feature point set in the blood vessel region as the second key feature point.

[0066] Step d: determining the degree of overlap between the line segment formed by the starting feature point and the second key feature point and the blood vessel area; if the degree of overlap is higher than a second threshold, adding the second key feature point to the venous feature data unit; if the degree of overlap is lower than the second threshold, excluding the second key feature point;

[0067] In step c of the embodiment of the present invention, after determining the starting feature point from the reference feature points, the second key feature points are determined one by one by traversing the feature points outside the reference feature point set one by one, and the line segments between the traversed second key feature points and the reference feature points are compared with the vascular coincidence degree of the vascular area by a threshold value, and the feature points that can represent the vascular information through the vectors formed by the line segments between the feature points are calculated, thereby forming a vascular feature data unit. The vascular feature data unit is also based on the captured eye image and the digitized eye and vascular information, thereby realizing accurate eye information processing and diagnosis in a computer-automated digitized form.

[0068] Step e, repeating steps cd until all feature points outside the reference feature point set of the vascular image are traversed based on the starting feature point, and the starting feature point and all second key feature points that meet the overlap degree higher than the second threshold are added to the venation feature data unit.

[0069] After the judgment between the first reference feature point and the first second key point in the combination of reference feature points in step c and step d is completed, all adjacent feature points outside the combination of reference feature points are traversed based on step c, and the second feature point that meets the rules of step d is added to the context feature data unit.

[0070] It should be noted that the methods in step c, step d, and step e in the embodiment of the present invention can achieve high efficiency and accuracy in extracting feature points that can accurately represent the vascular information in the eye vascular image among all feature points, and the overall demand for computing power is not large, and can be completed in computing units with higher performance, ensuring the low cost of eye diagnosis equipment and greater room for hardware selection.

[0071] In an embodiment of the present invention, the step of calculating the vector of each line segment formed by connecting the ordered reference feature points and key feature points in each context feature data unit and adding the vector to the context feature data unit further includes: determining a starting feature point from the reference feature points, and a line segment formed by connecting the key feature points in step d corresponding to the starting feature point and having a degree of overlap higher than a second threshold, generating a vector, and adding a plurality of the vectors to the context feature data unit.

[0072] Step f, repeating steps b-e until the traversal of the vein feature data units based on all M feature points is completed.

[0073] In the embodiment of the present invention, after all the second key feature points that meet the requirements of step d are calculated based on the first reference feature point in the reference feature point combination, the remaining M-1 reference feature points in the reference feature point combination and all the second key feature points that meet the requirements of step d are repeatedly traversed in the manner of steps b to e. It should be noted that, for example, the second key feature point that meets the requirements of step d and is associated with the first reference feature point in the reference feature point combination is not re-traversed in the extraction of all the second key feature points that meet the requirements of step d and are associated with the subsequent M-1 reference feature points. In other words, the second key feature point that meets the requirements of step d and is associated with the first reference feature point in the reference feature point combination is deleted from the feature point set outside the reference feature point set to be traversed during the subsequent traversal process.

[0074] In an embodiment of the present invention, the venation feature data unit is established by selecting a reference feature point set based on the distance from the edge of the blood vessel region, and then selecting multiple second key feature points that meet the requirements outside the reference feature point set based on the degree of overlap between the line segment connected by each reference feature point and the blood vessel region for each reference feature point in the reference feature point set. The multiple second key feature points corresponding to the reference feature points in the reference feature point set and the line segment vectors connected by the reference feature points and the second key feature points are used to form a venation feature data unit. The venation feature data unit formed in this manner can achieve high efficiency and accuracy in extracting feature points that can accurately represent vascular information in the ocular vascular image from all feature points, and the overall computing power requirement is not large. It can be completed in computing units with higher performance, ensuring the low cost of the eye analysis device and greater room for hardware selection. In addition, it can achieve good equipment automation and precise detection effects in the eye information processing process or when integrated into the eye analysis device, thereby realizing the expanded application of the device.

[0075] Attachment Figure 3 The ocular blood vessel feature point information processing device of the embodiment of the present invention shown in FIG. 1 corresponds to the ocular blood vessel feature point information processing device in embodiment 1, specifically including

[0076] A first extraction unit 301 is configured to obtain an input eye image and extract a blood vessel region from the eye image;

[0077] A second extraction unit 303 is used to extract feature points from the eye image;

[0078] The processing unit 305 establishes a vein feature data unit based on the degree of overlap between a line segment formed between at least two feature points and a blood vessel region of the eye image;

[0079] A first calculation unit 307 calculates a vector of each line segment formed by connecting the ordered reference feature points and key feature points in each venation feature data unit, and adds the vector to the venation feature data unit, wherein the reference feature points are feature points selected based on the distance from the edge of the blood vessel region; and the key feature points are feature points selected based on the degree of overlap between the line segment connected to the at least one reference feature point and the blood vessel region.

[0080] The second calculation unit 309 is configured to calculate the blood vessel area corresponding to each venous feature data unit by using a dilation algorithm based on a vector formed by connecting every two key feature points in the venous feature data unit.

[0081] In addition to the above-mentioned functions, each unit of the eye vein feature point information processing device in the embodiment of the present invention can also complete the steps, functions and effects of any implementation of the blood vessel feature point information processing method described in Example 1 of the present invention. Since any implementation of each specific step in the above Example 1 has been described in detail, it will not be repeated here. The correspondence between any implementation and the first extraction unit 301, the second extraction unit 303, the processing unit 305, the first calculation unit 307, and the second calculation unit 309 in the eye vein feature point information processing device shall be based on the description of each step in the vein feature point information processing method in Example 1 of the present invention.

[0082] Attachment Figure 4 A schematic diagram of an electronic device structure 400 of an eye vein feature point information processing device according to an embodiment of the present invention is shown as follows: Figure 4 As shown, the electronic device may include: a processor 410, a communication interface 420, a memory 430 and a communication bus 440, wherein the processor 410, the communication interface 420 and the memory 430 communicate with each other via the communication bus 440. The processor 410 may call the logic instructions in the memory 430 to execute a method of processing eye vein feature point information of the present invention, the method including:

[0083] Acquire an input eye image and extract a blood vessel region of the eye image;

[0084] extracting feature points from the eye image;

[0085] establishing a vein feature data unit based on a degree of overlap between a line segment formed between at least two feature points and a blood vessel region of the eye image;

[0086] Calculating the vector of each line segment formed by the ordered reference feature points and key feature points in each venation feature data unit, and adding the vector to the venation feature data unit, wherein the reference feature points are feature points selected and determined based on the distance from the edge of the blood vessel region; and the key feature points are feature points selected and determined based on the degree of overlap between the line segment formed by the at least one reference feature point and the blood vessel region;

[0087] Based on the vector formed by connecting every two key feature points in the venous feature data unit, a dilation algorithm is used to calculate the blood vessel area corresponding to each venous feature data unit.

[0088] In addition to processing the above steps, the method for executing instructions by the processor in the electronic device of the present invention can also complete the steps, functions and effects of any implementation of the blood vessel feature point information processing method described in Example 1 of the present invention. Since any implementation of each specific step in the above Example 1 has been described in detail, it will not be repeated here.

[0089] In addition, the logic instructions in the above-mentioned memory 430 can be implemented in the form of a software functional unit and can be stored in a computer-readable storage medium when sold or used as an independent product. Based on this understanding, the technical solution of the present invention, or the part that contributes to the prior art, or the part of the technical solution, can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to perform all or part of the steps of the method described in each embodiment of the present invention. The aforementioned storage medium includes: various media that can store program codes, such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a random access memory (RAM), a magnetic disk or an optical disk.

[0090] On the other hand, the present invention also provides a computer program product, which includes a computer program stored on a non-transitory computer-readable storage medium, and the computer program includes program instructions. When the program instructions are executed by a computer, the computer can perform a method of eye vein feature point information processing provided by the above methods, and the method includes:

[0091] Acquire an input eye image and extract a blood vessel region of the eye image;

[0092] extracting feature points from the eye image;

[0093] establishing a vein feature data unit based on a degree of overlap between a line segment formed between at least two feature points and a blood vessel region of the eye image;

[0094] Calculating the vector of each line segment formed by the ordered reference feature points and key feature points in each venation feature data unit, and adding the vector to the venation feature data unit, wherein the reference feature points are feature points selected and determined based on the distance from the edge of the blood vessel region; and the key feature points are feature points selected and determined based on the degree of overlap between the line segment formed by the at least one reference feature point and the blood vessel region;

[0095] Based on the vector formed by connecting every two key feature points in the venous feature data unit, a dilation algorithm is used to calculate the blood vessel area corresponding to each venous feature data unit.

[0096] On the other hand, the present invention also provides a non-transitory computer-readable storage medium having a computer program stored thereon, which is implemented when the computer program is executed by a processor to perform any of the above-mentioned eye vein feature point information processing methods disclosed herein.

[0097] The device embodiments described above are merely illustrative. The units described as separate components may or may not be physically separate, and the components shown as units may or may not be physical units, i.e., they may be located in one location or distributed across multiple network units. Some or all of the modules may be selected based on actual needs to achieve the objectives of the present embodiment. Persons of ordinary skill in the art will be able to understand and implement the present invention without inventive effort.

[0098] Through the description of the above embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by means of software plus a necessary general hardware platform, or of course, by hardware. Based on this understanding, the essence of the above technical solution or the part that contributes to the existing technology can be embodied in the form of a software product. The computer software product can be stored in a computer-readable storage medium, such as ROM / RAM, a magnetic disk, an optical disk, etc., and includes a number of instructions for enabling a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each embodiment or certain parts of the embodiments.

[0099] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit it. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the various embodiments of the present invention.

Claims

1. A method for processing eye vein feature point information, used for processing eye image information, characterized in that: The method comprises, Acquire an input eye image and extract a blood vessel region of the eye image; Extracting feature points from the eye image; Establishing a venation feature data unit based on the degree of overlap between a line segment formed between at least two feature points and a blood vessel region of the eye image; specifically comprising: Step a: extracting feature points in the eye image whose distance to the edge of the blood vessel region is less than a first preset threshold, and forming a reference feature point set with all M feature points whose distance to the edge of the blood vessel region is less than the first preset threshold, where M is a positive integer greater than or equal to 1; Step b: selecting the feature point with the shortest distance from the edge of the blood vessel region in the reference feature point set as the starting feature point; Step c: determining a feature point adjacent to the starting feature point outside the reference feature point set in the blood vessel region as a second key feature point; Step d, determining the degree of overlap between the line segment formed by the starting feature point and the second key feature point and the blood vessel region; if the degree of overlap is higher than a second threshold, adding the second key feature point to the venation feature data unit; if the degree of overlap is lower than the second threshold, excluding the second key feature point; Step e, repeating steps cd until all feature points outside the reference feature point set of the eye image are traversed based on the starting feature point, and the starting feature point and all second key feature points satisfying a degree of overlap higher than a second threshold are added to the venation feature data unit; Step f, repeating steps b-e until the traversal of the vein feature data units based on all M reference feature points is completed; Calculating the vector of each line segment formed by the ordered reference feature points and key feature points in each venation feature data unit, and adding the vector to the venation feature data unit, wherein the reference feature points are feature points selected and determined based on the distance from the edge of the blood vessel region; and the key feature points are feature points selected and determined based on the degree of overlap between the line segment formed by the at least one reference feature point and the blood vessel region; Based on the vector formed by connecting every two key feature points in the venous feature data unit, a dilation algorithm is used to calculate the blood vessel area corresponding to each venous feature data unit.

2. The information processing method according to claim 1, wherein: The extracting of the blood vessel region from the eye image specifically includes extracting the blood vessel region from the eye image using an adaptive binarization method.

3. The information processing method according to claim 1, wherein: The extracting of feature points from the eye image specifically includes extracting feature points from the blood vessel image using an ORB feature algorithm or a deep learning SuperPonit model.

4. The information processing method according to claim 1, wherein: The first preset threshold is the maximum distance between the 5% of feature points closest to the edge of the blood vessel region among all feature points and the edge of the blood vessel region.

5. The information processing method according to claim 1, wherein: The step of calculating the vector of each line segment formed by connecting the ordered reference feature points and key feature points in each context feature data unit and adding the vector to the context feature data unit further includes: determining a starting feature point from the reference feature points, and a line segment formed by connecting the key feature points in step d corresponding to the starting feature point and having a degree of overlap higher than a second threshold, generating a vector, and adding a plurality of the vectors to the context feature data unit.

6. An eye vein feature point information processing device, characterized in that: Specifically include: a first extraction unit, configured to acquire an input eye image and extract a blood vessel region of the eye image; a second extraction unit, configured to extract feature points from the eye image; a processing unit, which establishes a vein feature data unit based on a degree of overlap between a line segment formed between at least two feature points and a blood vessel region of the eye image; Specifically include: Step a: extracting feature points in the eye image whose distance to the edge of the blood vessel region is less than a first preset threshold, and forming a reference feature point set with all M feature points whose distance to the edge of the blood vessel region is less than the first preset threshold, where M is a positive integer greater than or equal to 1; Step b: selecting the feature point with the shortest distance from the edge of the blood vessel region in the reference feature point set as the starting feature point; Step c: determining a feature point adjacent to the starting feature point outside the reference feature point set in the blood vessel region as a second key feature point; Step d, determining the degree of overlap between the line segment formed by the starting feature point and the second key feature point and the blood vessel region; if the degree of overlap is higher than a second threshold, adding the second key feature point to the venation feature data unit; if the degree of overlap is lower than the second threshold, excluding the second key feature point; Step e, repeating steps cd until all feature points outside the reference feature point set of the eye image are traversed based on the starting feature point, and the starting feature point and all second key feature points satisfying a degree of overlap higher than a second threshold are added to the venation feature data unit; Step f, repeating steps b-e until the traversal of the vein feature data units based on all M reference feature points is completed; a first calculation unit, calculating a vector of each line segment formed by connecting the ordered reference feature points and key feature points in each venation feature data unit, and adding the vector to the venation feature data unit, wherein the reference feature points are feature points selected and determined based on the distance from the edge of the blood vessel region; and the key feature points are feature points selected and determined based on the degree of overlap between the line segment connected to the at least one reference feature point and the blood vessel region; The second calculation unit is configured to calculate the blood vessel area corresponding to each venous feature data unit by using a dilation algorithm based on a vector formed by connecting every two key feature points in the venous feature data unit.

7. A computer storage medium storing a plurality of instructions, wherein the instructions are suitable for being loaded by a processor and executing the steps of the information processing method according to any one of claims 1 to 5.

8. An electronic device comprising: A processor and a memory; wherein the memory stores a computer program, and the computer program is suitable for being loaded by the processor and executing the steps of the information processing method according to any one of claims 1 to 5.

Images