CN117115899A - Method and device for identifying white-eye venation characteristics, computer storage medium and electronic equipment - Google Patents

Abstract

The invention discloses a method, device, computer storage medium and electronic equipment for identifying the vein characteristics of the white eye based on the viscera partitioning method of the white eye of the human eye. The method includes: acquiring a human eye image to be recognized; inputting the human eye image to be recognized into a first network structure model, calculating feature point coordinate information of vein shape _, color and direction in the human eye image; The feature point coordinate information is classified and sorted to generate a feature point data unit, and the classification and sorting is based on the corresponding relationship between the feature point and the vein; the human eye vein characteristics corresponding to the feature point data unit are calculated through the second network structure model , the human eye vein characteristics include blood vessel image information of color, thickness, and spots; based on the feature point data unit and the human eye vein characteristics, a first image including at least the human eye vein characteristic information is generated. The invention provides a method for identifying the vein characteristics of the white of the human eye, which has the advantages of high recognition accuracy, small amount of adaptation data, rapid iteration in actual use, better versatility, and easy to be put into production quickly.

Description

A method, device, computer storage medium and electronic device for identifying the characteristics of white eye veins equipment

Technical field

The present invention relates to the field of human eye image recognition, and in particular to a method, device, computer storage medium and electronic equipment for identifying white eye vein characteristics.

Background technique

In recent years, based on eye auxiliary information processing, how to automatically and accurately extract effective features of eye images for comparison based on images has always been a difficulty in this field of research, and is also a major aspect that limits the application of eye analysis-related instruments and equipment. , How to extract traditional Chinese medicine context features from eye images is a common technical problem in the current field of eye analysis instrument and equipment identification technology. The current TCM eye information analysis system recognition method is based on deep learning of all pixels of the whole eye image to build an AI-based analysis model. Although it is relatively easy to implement, the image has too much irrelevant element information and cannot be directly processed without processing. Machine learning will cause great interference to the final results and lead to poor accuracy; at the same time, this model requires a large number of samples due to too many input elements, resulting in a long production cycle and poor stability. Due to the large number of samples , it takes a long time to re-learn machine learning after each sample modification, and it is impossible to respond quickly to problems that arise. Therefore, it is difficult to implement in practical applications and it is impossible to form a device that can be used commercially.

Therefore, how to provide a method and device that has good stability, high recognition accuracy, and can respond quickly and identify human eye vein characteristics is a technical problem that urgently needs to be broken through in this field.

Contents of the invention

Based on the above deficiencies in the prior art, the present invention provides a method for identifying the vein characteristics of the white eye based on the organ partitioning method of the white eye of the human eye. can improve the above shortcomings.

As one aspect of the present invention, the present invention provides a method for identifying the vein characteristics of the white eye based on the viscera partitioning method of the white eye of the human eye, wherein the identification method includes:

Obtain the human eye image to be recognized;

Input the human eye image to be recognized into the first network structure model, and calculate the feature point coordinate information of vein shape _, color and direction in the human eye image;

Classify and sort the feature point coordinate information to generate feature point data units, where the classification and sorting is based on the corresponding relationship between the feature points and the context;

The human eye vein characteristics corresponding to the feature point data unit are calculated through the second network structure model, and the human eye vein characteristics include blood vessel image information of color, thickness, and spots;

Based on the feature point data unit and human eye vein characteristics, a first image including at least human eye vein characteristic information is generated.

Preferably, in the method for identifying the characteristics of veins in the eyes provided by the first aspect of the present invention, the method further includes calculating a type of veins in the human eye based on the characteristics of the veins in the human eye, and the type of veins in the human eye is the same as that of human organs. There is corresponding relationship between vein shape, vein color and special vein information, and the human eye vein type includes partition information.

Preferably, the method for identifying the characteristics of veins in the white eye provided by the first aspect of the present invention is characterized in that the types of veins in the human eye at least include: thick roots, varicose, extended, disconnected, bifurcated, raised, blurred There are ten forms: a piece, a dew, a black circle, and a penetrating pupil.

and/or eight colors: bright red, purple red, dark red, red with black, red with yellow, light yellow, gray, and dark gray,

And/or there are six special veins: straight line, root deficiency, grid, dark spots, yellow spots, and livedo.

Preferably, in the method for identifying the characteristics of white eye veins provided by the first aspect of the present invention, the second network structure model is trained using a support vector machine algorithm, and the feature point data is calculated through the second network structure model. The human eye context feature corresponding to the unit specifically includes: inputting the feature point data unit into the second network structure model, and calculating the human eye context feature corresponding to the feature point data unit in the human eye image.

Preferably, the first aspect of the present invention provides a method for identifying the characteristics of white eye veins, and the first network structure model is trained by the following method:

Step a. Randomly generate one or more simple graphics. The simple graphics include one or more types of quadrilaterals, triangles, line segments, and cubes. Mark the vertices of the simple graphics as feature points, and compare the simple graphics with the corresponding The coordinates of the feature points are used as input to train the initial model using the MagicPoint algorithm.

Step b. Input the human eye image to be recognized into the initial model trained by the MagicPoint algorithm, conduct a second round of training, and obtain an advanced model.

Step c. Use the human eye image to be recognized and the feature point coordinates calculated by the advanced model based on the human eye image to be recognized as input, and use the MagicPoint algorithm to perform a third round of training to obtain a high-order model. ;

Step d. Taking the human eye image to be recognized as input, using the high-order model to calculate the coordinates of the feature points corresponding to the human eye image to be recognized;

Step e. Manually check the accuracy of the feature points extracted in step d. If accurate, proceed to g. If inaccurate, repeat step c to iterate the high-order model;

Step g. Use the human eye image to be recognized and the corresponding feature point coordinates as input, and use the SuperPoint algorithm to train and generate the final model.

Preferably, the first aspect of the present invention provides a method for identifying white eye vein characteristics, and the second network model is trained by the following steps:

Step A. Normalize each feature point data unit in the training set and convert it into a 4x32 two-dimensional matrix. Each column in the matrix corresponds to each line segment of the feature point data unit, and the 4 data in each column are feature point data in turn. The vector angle, vector length, width, and color of a unit line segment. A 32-line data unit can store up to 32 line segment data. The redundant data exceeding 32 line segments is discarded in the two-dimensional matrix, and less than 32 line segments are filled with 0s. ;

Step B. Receive human eye context features manually marked for feature point data units;

Step C: Use the complex feature point data unit data normalized in step A and the human eye vein type of each vein feature data unit marked in step B as input, and use the support vector machine algorithm to train the model.

Preferably, the first aspect of the present invention provides a method for identifying the characteristics of white eye veins. The method of obtaining the human eye image to be identified specifically includes:

Import the human eye image, which is captured by the camera,

Remove the background of the human eye image and retain the eye image,

Remove the highlight areas caused by reflected lighting sources on the human eye image, and use the expansion algorithm to repair the removed areas.

The Canny edge detection algorithm, SOBEL edge detection algorithm, and expansion algorithm are used to identify and crop the image, leaving only the white eye area.

The human eye image that retains only the white eye area after cropping is used as the human eye image to be recognized.

In another embodiment of the present invention, the present invention also provides a device for identifying the vein characteristics of the white eye based on the organ partitioning method of the white eye of the human eye, which specifically includes:

The acquisition unit acquires the human eye image to be recognized;

The first calculation unit inputs the human eye image to be recognized into the first network structure model and calculates the feature point coordinate information of the vein shape _, color and direction in the human eye image;

A first generation unit is configured to classify and sort the feature point coordinate information and generate a feature point data unit, where the classification and sorting is based on the corresponding relationship between the feature point and the context;

The second calculation unit calculates the human eye vein characteristics corresponding to the feature point data unit through the second network structure model. The human eye vein characteristics include blood vessel image information of color, thickness, and spots;

The second generation unit generates a first image including at least human eye vein feature information based on the feature point data unit and human eye vein characteristics.

In another embodiment of the present invention, a computer storage medium is also provided. The computer storage medium stores a plurality of instructions, and the instructions are suitable for being loaded by the processor and executed as any one of the above based on the white eyes and internal organs of the human eye. Method steps for identifying the characteristics of white eye veins by partitioning method. .

In another embodiment of the present invention, the present invention also provides an electronic device, including: a processor and a memory; wherein the memory stores a computer program, and the computer program is adapted to be loaded by the processor and executed as follows: Any of the above method steps are based on the viscera partitioning method of the white eye of the human eye.

The method for identifying the characteristics of the white eye veins based on the viscera partitioning method of the white eye of the human eye provided by the present invention has the advantages of high recognition accuracy, small amount of adaptation data, rapid iteration in actual use, better versatility, and easy to be put into production quickly. The application of analysis equipment provides a reliable and stable method for identifying the characteristics of human eye pulses and white eye veins.

The present invention decomposes the identification of the vein characteristics of the white eye of the human eye into the calculation of feature point coordinate information, the classification and sorting based on the corresponding relationship between the characteristic points and veins, the calculation of the human eye vein characteristics, and the classification and matching of the human eye vein characteristics. The information related to the changing human eye vein characteristics can be classified in an orderly manner. By dividing the calculation and identification into two models to complete different stages, the simplicity of the data is ensured, thereby achieving the stability of data processing. The modular idea also realizes the sample processing It is controllable, easy to adaptively update, fast to identify and easy to put into production. It solves the current problems of end-to-end deep learning for human eye image processing, such as low recognition accuracy, high sample number requirements, and slow sample training, resulting in long production and iteration cycles. It provides a basic guarantee for the accurate identification and widespread application of traditional Chinese medicine eye diagnosis.

Additional features and advantages of the invention will be set forth in the description which follows, and in part will be apparent from the description, or may be learned by practice of the invention. The objectives and other advantages of the invention may be realized and attained by the structure particularly pointed out in the written description and claims hereof as well as the appended drawings.

Description of drawings

The drawings described here are used to provide a further understanding of the present invention and constitute a part of the present invention. The illustrative 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 attached picture:

Figure 1 is a flow chart of a method for identifying vein characteristics of the white eye based on the viscera partitioning method of the white eye of the present invention according to Embodiment 1 of the present invention.

Figure 2 is a flow chart of the first mesh model training method in Embodiment 1 of the present invention.

Figure 3 is a flow chart of the second network model training method in Embodiment 1 of the present invention.

Figure 4 is a schematic diagram of the device for identifying vein characteristics of the white of human eyes in Embodiment 2 of the present invention.

Figure 5 is a schematic structural diagram of electronic equipment in Embodiment 3 of the present invention.

Detailed ways

The preferred embodiments of the present invention will be described below in conjunction with the accompanying drawings. It should be understood that the preferred embodiments described here are only used to illustrate and explain the present invention, and are not used to limit the present invention. In the case of no conflict, the present invention The embodiments and features of the embodiments may be combined with each other.

In view of the problems existing in the field of human eye image recognition such as large data interference, poor accuracy, poor stability, inability to timely adaptive update according to actual use, untimely response, and long production cycle, Implementation 1 of the present invention provides a method based on human eye white The white eye vein feature identification method based on the eye viscera partition method has the advantages of high recognition accuracy, small amount of adaptation data, rapid iteration in actual use, better versatility, and easy to put into production quickly. It provides a reliable solution for the application of traditional Chinese medicine eye analysis equipment. A stable method for identifying vein features in the white of human eyes.

Specifically, in Embodiment 1 of the present invention, a method for identifying the vein characteristics of the white eye based on the organ partitioning method of the white eye of the human eye specifically includes:

Step 101: Obtain the human eye image to be recognized,

Specifically, the human eye image can be an auxiliary support for fixing different people's eyes and presetting the human eye state through an auxiliary device, using a camera device to take pictures of the target area of the 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, it is preferred to use a camera that can clearly capture the details and colors around the eyes, including the eyes themselves, and a camera with corresponding high pixels and resolution, such as a SLR camera, a smartphone camera, or an industrial camera, etc. All are available as options.

The human eye image to be recognized in the embodiment of the present invention is the human eye image obtained by taking a picture with a camera and performing image data processing. Specifically, the obtaining the human eye image to be recognized specifically includes:

Import the human eye image, which is captured by the camera,

Remove the background of the human eye image and retain only the eye image,

Remove the highlight areas caused by reflected lighting sources on the human eye image, and use an improved expansion algorithm to repair the removed areas.

The Canny edge detection algorithm, SOBEL edge detection algorithm, and improved expansion algorithm are used to identify and crop the image, leaving only the white eye area.

The human eye image that retains only the white eye area after cropping is used as the human eye image to be recognized.

After the human eye image captured by the camera is processed through the above steps, it can be cropped to the surrounding area to focus on the eye area to be identified later, and at the same time, the image caused by factors such as reflected light, background, edge blur, etc. Eliminating influencing factors can provide high-quality human eye images for subsequent precise processing, thereby improving the accuracy of overall human eye vein feature recognition.

Step 102: Input the human eye image to be recognized into the first network structure model, and calculate the feature point coordinate information of vein shape _, color and direction in the human eye image.

Specifically, in the embodiment of the present invention, the feature points are points that can represent the shape characteristics of the image. For example, 4 vertices of a square, 3 vertices of a triangle, 8 vertices of a cube, the inflection point of a polyline, and the intersection of a cross line are specifically selected. , connect these points to get the outline of the graph. Since the shape and direction of human eye veins are related to the shape characteristics of the image, they can be accurately expressed using key points.

The feature point coordinate information calculated in Embodiment 1 of the present invention may be scattered feature point coordinates that do not correspond to the context state.

Specifically, as shown in Figure 2, the first mesh structure model in the embodiment of the present invention is trained by the following method:

Step a. Randomly generate one or more simple graphics. The simple graphics include one or more types of quadrilaterals, triangles, line segments, and cubes. Mark the vertices of the simple graphics as feature points, and compare the simple graphics with the corresponding The coordinates of the feature points are used as input to train the initial model using the MagicPoint algorithm.

Step b. Input the human eye image to be recognized into the initial model trained by the MagicPoint algorithm, conduct a second round of training, and obtain an advanced model.

Step c. Use the human eye image to be recognized and the feature point coordinates calculated by the advanced model based on the human eye image to be recognized as input, and use the MagicPoint algorithm to perform a third round of training to obtain a high-order model. ;

Step d. Taking the human eye image to be recognized as input, using the high-order model to calculate the coordinates of the feature points corresponding to the human eye image to be recognized;

Step e. Manually check the accuracy of the feature points extracted in step d. If accurate, proceed to g. If inaccurate, repeat step c to iterate the high-order model;

Step g. Use the human eye image to be recognized and the corresponding feature point coordinates as input, and use the SuperPoint algorithm to train and generate the final model.

The present invention sets a separate calculation model for feature point coordinate information, and quickly and accurately calculates and extracts the feature points in the human eye image information input to be recognized through the first step, and uses the feature point coordinate information for subsequent rapid and accurate processing. It provides a foundation. At the same time, the calculation process of feature points based on the human eye image to be recognized is compared with directly calculating the complete context features or context types. The data types and rules that need to be calculated are simple, the data volume is small, and it has the advantages of fast and accurate calculation.

Step 103: Classify and sort the feature point coordinate information to generate feature point data units, and the classification and sorting is based on the corresponding relationship between the feature points and the context.

It should be noted that in the embodiment of the present invention, the veins are human tissue information that can transmit human biological signals, including blood vessels and nerve patterns in the whites of human eyes, and are usually represented in the form of image information.

After calculating the feature point coordinate information that can characterize the vein shape and direction, the obtained feature point coordinate information is scattered feature point coordinates that are not associated with the corresponding vein shape and direction. The feature point coordinate information calculated from the first network structure model is scattered feature points and is not associated with the context. However, the scattered feature points actually represent the information contained in all contexts. Therefore, Step 103 sorts the scattered feature points according to the corresponding contexts through classification and sorting, which can achieve high efficiency in subsequent calculation of context features.

Therefore, step 103 in the embodiment of the present invention is used to further classify and sort based on the corresponding relationship between the feature points and the veins, filter out the feature points that can be associated with the veins, and exclude those that are not related to the actual blood vessel position and shape direction in the picture. These uncorrelated feature points are usually called miscalculated feature points in the process. After removing the miscalculated feature points, step 103 also completes sorting the feature points associated with the vein position and shape direction. , the sorting is divided into data groups according to the area where each context in the image to be identified is located, and each data group corresponds to the position of each context in the image to be identified.

In step 103 in the embodiment of the present invention, the classification and sorting of the feature point coordinate information can be performed by using existing technologies such as topological structures based on blood vessel graphics through adaptive graphics detectors, or by using big data. The trained machine learning is used for identification. The specific classification and sorting method can be selected and optimized by those skilled in the art based on the distribution rules of human eye blood vessels and nerves, as well as the calculated feature point coordinate set.

Step 104: Calculate the human eye vein characteristics corresponding to the feature point data unit through the second network structure model. The human eye vein characteristics include blood vessel image information of color, thickness, and spots. In step 104 of the embodiment of the present invention, the second network structure model is trained using a support vector machine algorithm, and the human eye context characteristics corresponding to the feature point data units are calculated through the second network structure model, which specifically includes: The feature point data unit inputs the second network structure model and calculates the human eye context characteristics corresponding to the feature point data unit in the human eye image.

Specifically, as shown in Figure 2, the second network model in the present invention is trained by the following steps:

Step A. Normalize each feature point data unit in the training set and convert it into a 4x32 two-dimensional matrix. Each column in the matrix corresponds to each line segment of the feature point data unit, and the 4 data in each column are feature point data in turn. The vector angle, vector length, width, and color of a unit line segment. A 32-line data unit can store up to 32 line segment data. The redundant data exceeding 32 line segments is discarded in the two-dimensional matrix, and less than 32 line segments are filled with 0s. ;

Step B. Receive human eye context features manually marked for feature point data units;

Step C: Use the complex feature point data unit data normalized in step A and the human eye vein type of each vein feature data unit marked in step B as input, and use the support vector machine algorithm to train the model.

The human eye vein characteristics in the embodiment of the present invention at least include 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 veins of the human eye can be included in the present invention.

In Embodiment 1 of the present invention, through steps 102 to 104, the human eye image to be recognized is processed by models and algorithms to obtain human eye venation features that at least include blood vessel and nerve information. The human eye venation features include the shape of human eye blood vessels, Characteristics such as direction, color, and thickness can reflect the distribution of blood vessels in the human eye, and at the same time, in order to be able to correlate and display different states of the human eye, specifically, the types of human eye veins described in the present invention include at least: thick roots, varicose veins, extensions, There are ten forms: disconnection, bifurcation, bulge, blur, dew, black circle, and penetrating pupils.

and/or eight colors: bright red, purple red, dark red, red with black, red with yellow, light yellow, gray, and dark gray,

And/or there are six special veins: straight line, root deficiency, grid, dark spots, yellow spots, and livedo.

Step 105: Generate a first image including at least human eye vein feature information based on the feature point data unit and human eye vein characteristics.

Based on the classification theory of human eye status based on traditional Chinese medicine, and based on the corresponding relationship between human eye vein characteristics and human eye vein types, the human eye vein characteristics are obtained into human eye vein types. The human eye image information after automatic calculation by the machine is converted into human eye status information. It also provides a guarantee for automatically providing analysis of human eye status. It has achieved breakthroughs in machine automation, high efficiency and accuracy, and has played a very important role in the application and promotion of eye diagnostic instruments.

It should be noted that the division method of the internal organs of the white eye of the human eye comes from the eye acupuncture theory of Professor Peng Jingshan, a famous veteran of traditional Chinese medicine. It is part of the national key basic research and development plan (973 plan) in the field of traditional Chinese medicine, project number 2007CB512707, through the left and right eyes. The theory of eight zones and thirteen acupoints is used to identify eye information through eye divisions, vein characteristics and the correspondence between internal organs. The method for identifying the vein characteristics of the white eye in the embodiment of the present invention is based on the application of the viscera partitioning method of the white eye of the human eye in computer image processing.

Embodiment 1 of the present invention provides a method for identifying vein characteristics of the white eye of the human eye, which has the advantages of high recognition accuracy, adaptable data volume and rapid iteration in actual use, better versatility, and ease of rapid production, and is widely used in the field of traditional Chinese medicine eye diagnosis. Provides a reliable and stable method for identifying the vein characteristics of the white of human eyes.

The present invention decomposes the identification of the vein characteristics of the white eye of the human eye into the calculation of feature point coordinate information, the classification and sorting based on the corresponding relationship between the characteristic points and veins, the calculation of the human eye vein characteristics, and the classification and matching of the human eye vein characteristics. The information related to the changing human eye vein characteristics can be classified in an orderly manner. By dividing the calculation and identification at different stages into two models, the single accuracy of the data is ensured, thereby achieving the stability of data processing, and the modular idea is also realized. Samples are controllable, easy to adaptively update, fast to identify and easy to put into production. It solves the current problems of end-to-end deep learning in human eye image processing, such as low recognition accuracy, high sample number requirements, and slow sample training, which lead to long production and iteration cycles. This provides a basic guarantee for the accurate identification and widespread application of traditional Chinese medicine eye diagnosis.

As shown in Figure 4, the second embodiment of the present invention provides a device for identifying the vein type of the white of the human eye, which specifically includes an acquisition unit 401 to acquire the human eye image to be recognized;

The first calculation unit 403 inputs the human eye image to be recognized into the first network structure model, and calculates the feature point coordinate information of the vein shape _, color and direction in the human eye image;

The first generation unit 405 classifies and sorts the feature point coordinate information and generates a feature point data unit, where the classification and sorting is based on the corresponding relationship between the feature point and the context;

The second calculation unit 407 calculates the human eye vein characteristics corresponding to the feature point data unit through the second network structure model. The human eye vein characteristics include blood vessel image information of color, thickness, and spots;

The second generation unit 409 generates a first image including at least human eye context feature information based on the feature point data unit and human eye context characteristics.

The human eye image to be recognized in the embodiment of the present invention is the human eye image obtained by taking a picture with a camera and performing image data processing. Specifically, the obtaining the human eye image to be recognized specifically includes:

Import the human eye image, which is captured by the camera,

Remove the background of the human eye image and retain only the eye image,

Remove the highlight areas caused by reflected lighting sources on the human eye image, and use an improved expansion algorithm to repair the removed areas.

The Canny edge detection algorithm, SOBEL edge detection algorithm, and improved expansion algorithm are used to identify and crop the image, leaving only the white eye area.

The human eye image that retains only the white eye area after cropping is used as the human eye image to be recognized.

After the human eye image captured by the camera is processed through the above steps, it can be cropped to the surrounding area to focus on the eye area to be identified later, and at the same time, the image caused by factors such as reflected light, background, edge blur, etc. The elimination of influencing factors can provide high-quality human eye images for subsequent precise processing, thereby improving the accuracy of the overall human eye white eye vein feature recognition.

The composition of the first network structure model and the second network structure model and the training method in the embodiment of the present invention are the contents that have been fully disclosed in Embodiment 1, and will not be described again here.

A device for identifying the vein characteristics of the white of the human eye in Embodiment 2 of the present invention shown in Figure 4 corresponds to the method for identifying the vein characteristics of the white of the human eye in Embodiment 1, and is calculated by decomposing the vein characteristic recognition of the white of the human eye into feature point coordinate information. , based on the classification and sorting of the corresponding relationship between feature points and veins, as well as the calculation of human eye vein characteristics, the classification and matching of human eye vein characteristics, the complex and changeable human eye vein feature related information can be classified in an orderly manner, by dividing it into two Each model completes different stages of calculation and identification respectively, ensuring the simplicity of the data, thereby achieving the stability of data processing. The modular idea also achieves controllable samples, facilitates adaptive updates, rapid identification, and facilitates production, solving the current problem. End-to-end deep learning in human eye image processing has low recognition accuracy, high sample number requirements, and slow sample training, resulting in long production and iteration cycles. It provides innovative support for the accurate recognition and widespread application of traditional Chinese medicine eye diagnosis.

Figure 5 illustrates a schematic structural diagram of an electronic device in Embodiment 3 of the present invention. As shown in Figure 5, the electronic device may include: a processor (processor) 510, a communications interface (Communications Interface) 520, and a memory (memory) 530 and the communication bus 540, wherein the processor 510, the communication interface 520, and the memory 530 complete communication with each other through the communication bus 540. The processor 510 can call the logical instructions in the memory 530 to execute the present invention's white eye vein feature recognition method based on the white eye viscera partitioning method. The method includes: acquiring the human eye image to be recognized;

Input the human eye image to be recognized into the first network structure model, and calculate the feature point coordinate information of vein shape _, color and direction in the human eye image;

Classify and sort the feature point coordinate information to generate feature point data units, where the classification and sorting is based on the corresponding relationship between the feature points and the context;

The human eye vein characteristics corresponding to the feature point data unit are calculated through the second network structure model, and the human eye vein characteristics include blood vessel image information of color, thickness, and spots;

Based on the feature point data unit and human eye vein characteristics, a first image including at least human eye vein characteristic information is generated.

In addition, the above-mentioned logical instructions in the memory 530 can be implemented in the form of software functional units 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 essentially or the part that contributes to the existing technology 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, including Several instructions are used to cause a computer device (which may be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the methods described in various embodiments of the present invention. The aforementioned storage media include: U disk, mobile hard disk, read-only memory (ROM, Read-Only Memory), random access memory (RAM, Random Access Memory), magnetic disk or optical disk and other media that can store program code. .

On the other hand, the present invention also provides a computer program product, the computer program product includes a computer program stored on a non-transitory computer-readable storage medium, the computer program includes program instructions, when the program instructions are executed by a computer , the computer can execute the white eye vein feature recognition method based on the viscera partitioning method of the white eye of the human eye provided in the above embodiments, the method includes: obtaining the human eye image to be recognized;

Input the human eye image to be recognized into the first network structure model, and calculate the feature point coordinate information of vein shape, color and direction in the human eye image;

Classify and sort the feature point coordinate information to generate feature point data units, where the classification and sorting is based on the corresponding relationship between the feature points and the context;

The human eye vein characteristics corresponding to the feature point data unit are calculated through the second network structure model, and the human eye vein characteristics include blood vessel image information of color, thickness, and spots;

Based on the feature point data unit and human eye vein characteristics, a first image including at least human eye vein characteristic information is generated.

On the other hand, the present invention also provides a non-transitory computer-readable storage medium on which a computer program is stored. The computer program is implemented when executed by a processor to execute the above-mentioned one based on the human eye white eye viscera partitioning method. A method for identifying the characteristics of white eye veins, which method includes: obtaining a human eye image to be recognized;

Input the human eye image to be recognized into the first network structure model, and calculate the feature point coordinate information of vein shape _, color and direction in the human eye image;

Classify and sort the feature point coordinate information to generate feature point data units, where the classification and sorting is based on the corresponding relationship between the feature points and the context;

The human eye vein characteristics corresponding to the feature point data unit are calculated through the second network structure model, and the human eye vein characteristics include blood vessel image information of color, thickness, and spots;

Based on the feature point data unit and human eye vein characteristics, a first image including at least human eye vein characteristic information is generated.

The device embodiments described above are only illustrative. The units described as separate components may or may not be physically separated. The components shown as units may or may not be physical units, that is, they may be located in One location, or it can be distributed across multiple network units. Some or all of the modules can be selected according to actual needs to achieve the purpose of the solution of this embodiment. Persons of ordinary skill in the art can understand and implement the method without any creative effort.

Through the above description of the embodiments, those skilled in the art can clearly understand that each embodiment can be implemented by software plus a necessary general hardware platform, and of course, it can also be implemented by hardware. Based on this understanding, the part of the above technical solution that essentially 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, magnetic disk, optical disk, etc., including a number of instructions to cause a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in various embodiments or certain parts of the embodiments.

Finally, it should be noted that the above embodiments are only used to illustrate the technical solution of the present invention, but not to limit it; although the present invention has been described in detail with reference to the foregoing embodiments, those of ordinary skill in the art should understand that it can still be used Modifications are made to the technical solutions described in the foregoing embodiments, or equivalent substitutions are made to some of the technical features; however, these modifications or substitutions do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims (10)

1. A method for identifying the characteristics of the veins of the white eye based on the viscera partitioning method of the white eye of the human eye, characterized in that the identification method includes: Obtain the human eye image to be recognized; Input the human eye image to be recognized into the first network structure model, and calculate the feature point coordinate information of vein shape _, color and direction in the human eye image; Classify and sort the feature point coordinate information to generate feature point data units, where the classification and sorting is based on the corresponding relationship between the feature points and the context; The human eye vein characteristics corresponding to the feature point data unit are calculated through the second network structure model, and the human eye vein characteristics include blood vessel image information of color, thickness, and spots; Based on the feature point data unit and human eye vein characteristics, a first image including at least human eye vein characteristic information is generated. 2. The white eye vein feature identification method as claimed in claim 1, characterized in that the method further includes calculating a human eye vein type based on the human eye vein characteristics, and the human eye vein type is related to the human viscera type. There is corresponding relationship between vein shape, vein color and special vein information, and the human eye vein type includes partition information. 3. The method for identifying the characteristics of veins in white eyes as claimed in claim 2, characterized in that the types of veins in human eyes at least include: thick roots, varicose veins, extension, disconnection, bifurcation, ridge, blur, and dew. , black circle, penetrating pupil ten shapes; and/or eight colors of bright red, purple red, deep red, red with black, red with yellow, light yellow, gray, dark gray; and/or straight line, root void, net There are six special veins: grid, dark spots, yellow spots and blue spots. 4. The white eye vein feature identification method as claimed in claim 1, characterized in that the second network structure model is trained using a support vector machine algorithm, and the feature point data is calculated through the second network structure model. The human eye context feature corresponding to the unit specifically includes: inputting the feature point data unit into the second network structure model, and calculating the human eye context feature corresponding to the feature point data unit in the human eye image. 5. The white eye vein feature identification method as claimed in claim 1, characterized in that the first network structure model is trained by the following method: Step a. Randomly generate one or more simple graphics. The simple graphics include one or more types of quadrilaterals, triangles, line segments, and cubes. Mark the vertices of the simple graphics as feature points, and compare the simple graphics with the corresponding The coordinates of the feature points are used as input to train the initial model using the MagicPoint algorithm. Step b. Input the human eye image to be recognized into the initial model trained by the MagicPoint algorithm, conduct a second round of training, and obtain an advanced model. Step c. Use the human eye image to be recognized and the feature point coordinates calculated by the advanced model based on the human eye image to be recognized as input, and use the MagicPoint algorithm to perform a third round of training to obtain a high-order model. ; Step d. Taking the human eye image to be recognized as input, using the high-order model to calculate the coordinates of the feature points corresponding to the human eye image to be recognized; Step e. Manually check the accuracy of the feature points extracted in step d. If accurate, proceed to g. If inaccurate, repeat step c to iterate the high-order model; Step g. Use the human eye image to be recognized and the corresponding feature point coordinates as input, and use the SuperPoint algorithm to train and generate the final model. 6. The white eye vein feature recognition method as claimed in claim 1, characterized in that the second network model is trained by the following steps: Step A. Normalize each feature point data unit in the training set and convert it into a 4x32 two-dimensional matrix. Each column in the matrix corresponds to each line segment of the feature point data unit, and the 4 data in each column are feature point data in turn. The vector angle, vector length, width, and color of a unit line segment. The redundant data of more than 32 line segments are discarded in the two-dimensional matrix, and less than 32 line segments are filled with 0; Step B. Receive human eye context features manually marked for feature point data units; Step C: Use the complex feature point data unit data normalized in step A and the human eye vein type of each vein feature data unit marked in step B as input, and use the support vector machine algorithm to train the model. 7. The method for identifying the characteristics of white eye veins as claimed in claim 1, characterized in that said obtaining the human eye image to be identified specifically includes: Import the human eye image, which is captured by the camera, Remove the background of the human eye image and retain the eye image, Remove the highlight areas caused by reflected lighting sources on the human eye image, and use the expansion algorithm to repair the removed areas. The Canny edge detection algorithm, SOBEL edge detection algorithm, and expansion algorithm are used to identify and crop the image, leaving only the white eye area. The human eye image that retains only the white eye area after cropping is used as the human eye image to be recognized. 8. A device for identifying the characteristics of the veins of the white eye based on the viscera partitioning method of the white eye of the human eye, which is characterized in that it specifically includes: The acquisition unit acquires the human eye image to be recognized; The first calculation unit inputs the human eye image to be recognized into the first network structure model and calculates the feature point coordinate information of the vein shape _, color and direction in the human eye image; A first generation unit is configured to classify and sort the feature point coordinate information and generate a feature point data unit, where the classification and sorting is based on the corresponding relationship between the feature point and the context; The second calculation unit calculates the human eye vein characteristics corresponding to the feature point data unit through the second network structure model. The human eye vein characteristics include blood vessel image information of color, thickness, and spots; The second generation unit generates a first image including at least human eye vein feature information based on the feature point data unit and human eye vein characteristics. 9. A computer storage medium, the computer storage medium stores a plurality of instructions, the instructions are suitable for the processor to load and execute the white eye meridian based on the human eye white eye viscera partitioning method according to any one of claims 1 to 7. Feature identification method steps. 10. An electronic device, comprising: a processor and a memory; wherein the memory stores a computer program, and the computer program is adapted to be loaded by the processor and executed as claimed in any one of claims 1-7 based on the white eyes of the human eye and the internal organs of the human body. Method steps for identifying the characteristics of white eye veins by partitioning method.