A Tongue Image-Assisted Recognition Method, System and Medium for Diabetic Nephropathy Patients

Abstract

The present invention relates to the field of image processing technology, and particularly relates to a method, system and medium for assisting in the recognition of tongue images of patients with diabetic nephropathy. The method includes: obtaining a tongue image region, clustering the tongue image region based on a preset number of different clustering clusters, and screening the optimal clustering result according to the clustering result; combining the gray-scale gradient rules of pixel points in different directions within the neighborhood of each pixel point and the relative position relationship between the pixel point and each cluster in the optimal clustering result, to obtain the belonging degree of each pixel point to each cluster in the optimal clustering result; according to the difference in the belonging degree of each pixel point and its neighborhood pixel points to the same cluster in the optimal clustering result, obtaining the belonging performance degree of each pixel point to each cluster in the optimal clustering result; and re-dividing the tongue image region according to the belonging performance degree and the initial membership degree of each pixel point to each cluster in the optimal clustering result. The present invention improves the accuracy of the division result of the tongue image region.

Description

Technical Field

[0001] The present invention relates to the technical field of image processing, and particularly to a method, system and medium for assisting in the recognition of tongue images of patients with diabetic nephropathy. Background Art

[0002] Diabetic nephropathy affects the quality of life of patients. As a non-invasive biomarker, tongue images can reflect the internal physiological and pathological states of the human body. However, traditional tongue image recognition methods mainly rely on doctors' experience and subjective judgment, and doctors need to repeatedly and carefully observe each detail, resulting in a large workload for doctors.

[0003] During the development of the disease in patients with diabetic nephropathy, internal changes in the human body will be reflected on the tongue body, showing color changes and texture hyperplasia. By using image processing to assist tongue diagnosis, clustering algorithms are usually used to divide the tongue image area, such as segmenting and quantifying areas with abnormal tongue coating colors, ecchymosis and cracks, etc., as an auxiliary standard for identifying the development stage of diabetic nephropathy. Therefore, the image of the tongue area is often clustered to extract features related to diabetic nephropathy, thereby reducing the workload of doctors. However, areas with obvious abnormal tongue coating colors are easy to segment, and other abnormalities are not well clustered during the FCM clustering process because their ranges in the entire tongue body distribution space are small and this part will be covered by adjacent areas. Therefore, the clustering effect is not good, which in turn affects the subsequent analysis results. Summary of the Invention

[0004] In order to solve the problem of poor clustering effect when clustering pixel points in the tongue image area by the existing method, the purpose of the present invention is to provide a method, system and medium for assisting in the recognition of tongue images of patients with diabetic nephropathy. The specific technical solutions adopted are as follows:

[0005] In a first aspect, the present invention provides a method for assisting in the recognition of tongue images of patients with diabetic nephropathy, the method comprising the following steps:

[0006] Obtain a tongue image area of a patient with diabetic nephropathy to be detected;

[0007] Based on different preset numbers of clustering clusters, cluster the pixel points in the tongue image area respectively, and screen the optimal clustering result according to the gray-scale dispersion of the pixel points in each clustering result; combine the gray-scale gradient law of the pixel points in different directions within the neighborhood of each pixel point in the tongue image area and the relative position relationship between the pixel point and each cluster in the optimal clustering result to obtain the membership degree of each pixel point in each cluster in the optimal clustering result; according to the difference between the membership degree of each pixel point and its neighborhood pixel points in the same cluster in the optimal clustering result, obtain the membership performance degree of each pixel point and each cluster in the optimal clustering result.

[0008] According to the membership manifestation degree and the initial membership degree of each pixel point to each cluster in the optimal clustering result, the tongue image region is re-divided to obtain different regions.

[0009] Preferably, the screening of the optimal clustering result according to the gray-scale dispersion of pixel points in each clustering result includes:

[0010] Calculating the gray-scale dispersion degree of all pixel points in each cluster in the candidate result, and obtaining the gray-scale consistency index of the candidate result according to the gray-scale dispersion degree of all pixel points in all clusters in the candidate result. The gray-scale dispersion degree is negatively correlated with the gray-scale consistency index; according to the gray-scale consistency index of the candidate result and the dispersion degree of the average gray-scale value of pixel points within all clusters, the segmentation evaluation value of the candidate result is obtained; the candidate result is any one of the clustering results;

[0011] Taking the clustering result corresponding to the maximum segmentation evaluation value as the optimal clustering result.

[0012] Preferably, the obtaining of the membership degree of each pixel point to each cluster in the optimal clustering result by combining the gray-scale gradual change law of pixel points in different directions within the neighborhood of each pixel point in the tongue image region and the relative position relationship between the pixel point and each cluster in the optimal clustering result includes:

[0013] Performing curve fitting on the gray-scale values of pixel points within the neighborhood of the candidate pixel point in the feature direction and the gray-scale value of the candidate pixel point respectively to obtain the fitting curve corresponding to the candidate pixel point in the feature direction;

[0014] According to the slope difference between each pixel point on the fitting curve and its adjacent pixel point, obtaining the gray-scale gradual change consistency of the candidate pixel point in the feature direction. The slope difference is negatively correlated with the gray-scale gradual change consistency;

[0015] Determining the direction from the center point of each cluster in the optimal clustering result to the candidate pixel point as the gray-scale change direction of the candidate pixel point relative to each cluster;

[0016] Obtaining the membership degree of the candidate pixel point to each cluster in the optimal clustering result according to the gray-scale gradual change consistency of the candidate pixel point in the gray-scale change direction relative to each cluster;

[0017] The candidate pixel point is any pixel point in the tongue image region, and the feature direction is any direction.

[0018] Preferably, the obtaining of the membership degree of the candidate pixel point to each cluster in the optimal clustering result according to the gray-scale gradual change consistency of the candidate pixel point in the gray-scale change direction relative to each cluster includes:

[0019] Obtain the membership degree of a candidate pixel point in the cluster to be analyzed according to the difference between the gray-scale gradient consistency of the candidate pixel point in the gray-scale change direction relative to the cluster to be analyzed and the gray-scale gradient consistency of the gray-scale change direction of other clusters except the cluster to be analyzed, and the difference between the gray-scale gradient consistencies is negatively correlated with the membership degree;

[0020] The cluster to be analyzed is any one of the optimal clustering results.

[0021] Preferably, obtaining the membership performance of each pixel point with each cluster in the optimal clustering result according to the difference between the membership degrees of each pixel point and its neighboring pixel points in the same cluster in the optimal clustering result includes:

[0022] Obtain the membership performance of the candidate pixel point with the cluster to be analyzed according to the difference between the membership degree of the candidate pixel point in the cluster to be analyzed and the membership degree of the pixel points within the neighborhood of the candidate pixel point in the cluster to be analyzed, and the difference between the membership degrees is negatively correlated with the membership performance.

[0023] Preferably, re-dividing the tongue image region to obtain different regions according to the membership performance and the initial membership degree of each pixel point with each cluster in the optimal clustering result includes:

[0024] For any pixel point in the tongue image region: take the product of the membership performance of the pixel point with the cluster to be analyzed and the initial membership degree of the pixel point with the cluster to be analyzed as the final membership degree of the pixel point with the cluster to be analyzed;

[0025] Divide each pixel point into the cluster with the largest final membership degree respectively to obtain different regions.

[0026] Preferably, use the FCM clustering algorithm to cluster all pixel points in the tongue image region.

[0027] Preferably, when clustering the pixel points in the tongue image region based on different preset numbers of clustering clusters respectively, it further includes: obtaining the initial membership degree of each pixel point in the tongue image region with each cluster in the clustering result.

[0028] In a second aspect, the present invention provides a tongue image assisted recognition system for diabetic nephropathy patients, and the system includes:

[0029] An image acquisition module, configured to acquire a tongue image region of a diabetic nephropathy patient to be detected;

[0030] The attribution expression calculation module is used to cluster the pixel points in the tongue image region based on different preset numbers of clustering clusters respectively, and screen the optimal clustering result according to the gray-scale dispersion of the pixel points in each clustering result; combining the gray-scale gradient law of the pixel points in different directions within the neighborhood of each pixel point in the tongue image region and the relative position relationship between the pixel point and each cluster in the optimal clustering result, obtain the attribution degree of each pixel point to each cluster in the optimal clustering result; according to the difference between the attribution degree of each pixel point and its neighborhood pixel points to the same cluster in the optimal clustering result, obtain the attribution expression degree of each pixel point to each cluster in the optimal clustering result.

[0031] The region division module is used to re-divide the tongue image region according to the attribution expression degree and the initial membership degree of each pixel point to each cluster in the optimal clustering result to obtain different regions.

[0032] In a third aspect, the present invention provides a medium that stores at least one computer-executable program. When the at least one program is run by a computer, the computer is caused to execute the steps in the above-mentioned tongue image-assisted recognition method for diabetic nephropathy patients.

[0033] The present invention has at least the following beneficial effects:

[0034] The present invention first collects the tongue image region of the to-be-detected diabetic nephropathy patient, then sets different numbers of clustering clusters respectively, and performs multiple clusterings on the pixel points in the tongue image region, and analyzes and screens the optimal clustering result according to the gray-scale dispersion of the pixel points in each clustering result; for diabetic nephropathy patients, their tongue images will show special manifestations, and the pixel points in the tongue image region will show a certain gray-scale change law. Further, combining the gray-scale gradient law of the pixel points in different directions within the neighborhood of each pixel point in the tongue image region and the relative position relationship between the pixel point and each cluster in the optimal clustering result, adjusts the initial membership degree of each pixel point to each cluster in the optimal clustering result, so that the pixel points in the tongue image region are not only divided according to color differences, but also combined with the special texture change law of the tongue image of diabetic nephropathy patients, making the division of each pixel point more accurate and the division result of the region more complete and accurate. Description of the Drawings

[0035] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present invention or the prior art, the following will briefly introduce the drawings required to be used in the description of the embodiments or the prior art. Obviously, the following drawings are only some embodiments of the present invention. For those of ordinary skill in the art, without creative efforts, other drawings can also be obtained according to these drawings.

[0036] Figure 1Flowchart of a method for assisting in the identification of tongue images of patients with diabetic nephropathy provided by an embodiment of the present invention;

[0037] Figure 2 Optimal clustering result diagram of the tongue image region provided by an embodiment of the present invention;

[0038] Figure 3 Flowchart of a method for obtaining the attribution expression degree provided by an embodiment of the present invention;

[0039] Figure 4 Structural block diagram of a system for assisting in the identification of tongue images of patients with diabetic nephropathy provided by an embodiment of the present invention. Detailed implementation manners

[0040] In order to further elaborate on the technical means and effects adopted by the present invention to achieve the predetermined invention purpose, the following describes in detail a method, a system, and a medium for assisting in the identification of tongue images of patients with diabetic nephropathy proposed according to the present invention in conjunction with the accompanying drawings and preferred embodiments.

[0041] Unless otherwise defined, all technical and scientific terms used herein have the same meaning as commonly understood by those skilled in the technical field to which the present invention belongs.

[0042] The following specifically describes the specific solutions of a method, a system, and a medium for assisting in the identification of tongue images of patients with diabetic nephropathy provided by the present invention in conjunction with the accompanying drawings.

[0043] An embodiment of a method for assisting in the identification of tongue images of patients with diabetic nephropathy:

[0044] The specific scenario targeted by this embodiment: When analyzing the symptoms of patients with diabetic nephropathy, first collect the tongue image region of patients with diabetic nephropathy, then set different clustering parameters, cluster the pixel points in the tongue image region of patients with diabetic nephropathy, and select the optimal clustering result. Furthermore, based on the optimal clustering result, the pixel points are divided again to improve the accuracy of the division result of the pixel points in the tongue image region, facilitating subsequent analysis.

[0045] This embodiment proposes a method for assisting in the identification of tongue images of patients with diabetic nephropathy. As Figure 1 shown, a method for assisting in the identification of tongue images of patients with diabetic nephropathy in this embodiment includes the following steps:

[0046] Step S1, obtain the tongue image region of the to-be-detected patient with diabetic nephropathy.

[0047] In the early morning fasting state, the diabetic nephropathy patient to be tested cleans the oral cavity with clean water, disinfects the outer ring of the sampling instrument with an alcohol cotton ball. The diabetic nephropathy patient to be tested takes a sitting position, places the mandible on the tongue surface fixing rack, fully fits the face with the fixing frame, the muscles are in a relaxed and natural state, the mouth is naturally opened wide, the tongue body is naturally extended out of the mouth, the tongue surface is flat and not curled, the tip of the tongue is naturally downward, and this fixed posture is maintained for several seconds to complete the image capture of the tongue image area. Perform operations such as filtering and denoising, color correction, and grayscale processing on the collected tongue image. Denote the image processed at this time as the tongue image area of the diabetic nephropathy patient to be tested for subsequent analysis and processing. Filtering and denoising, color correction, and grayscale processing are all existing technologies and will not be elaborated here.

[0048] Thus, the tongue image area of the diabetic nephropathy patient to be tested is obtained.

[0049] Step S2: Cluster the pixel points in the tongue image area based on different preset numbers of clustering clusters respectively, and screen the optimal clustering result according to the gray-scale dispersion of the pixel points in each clustering result; combine the gray-scale gradient law of the pixel points in different directions within the neighborhood of each pixel point in the tongue image area and the relative position relationship between the pixel point and each cluster in the optimal clustering result to obtain the membership degree of each pixel point in each cluster in the optimal clustering result; according to the difference between the membership degree of each pixel point and its neighborhood pixel points in the same cluster in the optimal clustering result, obtain the membership performance degree of each pixel point and each cluster in the optimal clustering result.

[0050] The tongue image is one of the important indicators reflecting the visceral functions and health status of the human body. Diabetic nephropathy is a common complication of diabetic patients. During its development process, the body functions will change accordingly and be reflected in the tongue image, mainly manifested as changes in the color of the tongue body and tongue coating, as well as the appearance of abnormal manifestations such as ecchymosis and cracks. In the process of doctors diagnosing diseases through tongue images, relying on the subjective experience of doctors, the light conditions or weak abnormal symptoms may all lead to missed diagnoses. While images can objectively record and save tongue images, be repeatedly viewed for various operations, and observe the details of tongue images more carefully. At the same time, through matching a large number of cases, it can provide a more comprehensive and accurate basis for doctors' diagnosis and treatment.

[0051] In the process of using tongue image to assist in the identification of diabetic nephropathy, simulate the actual tongue diagnosis steps of doctors. First, observe the color of the tongue body and tongue coating, and secondly, observe abnormal manifestations such as cracks. Different regions of the tongue image of disease patients have different manifestations. To better observe and analyze the tongue image, segment these regions in the tongue image to assist doctors in analysis.

[0052] When analyzing the tongue image manifestations of patients with diabetes nephropathy to be detected, this embodiment first clusters the pixel points in the tongue image region of the patients with diabetes nephropathy to be detected, and then analyzes the gray-scale change law of the pixel points in each cluster in combination with the clustering results to obtain the possibility of the pixel points belonging to the region, confirm their membership, adjust the membership output by clustering, and re-divide the pixel points, so as to improve the accuracy of the tongue image segmentation region and reduce errors.

[0053] Step S21: Cluster the pixel points in the tongue image region multiple times based on different preset numbers of clustering clusters, and screen the optimal clustering result according to the gray-scale dispersion of the pixel points in each clustering result.

[0054] The boundaries between the regions of the tongue image are generally blurred. For example, the tongue coating and the tongue body gradually transition. The FCM clustering algorithm can better handle the fuzziness and more flexibly segment the regions. Therefore, the FCM clustering algorithm is used to segment the tongue image region with fuzzy boundary features.

[0055] The tongue image of normal people shows the color difference between the tongue coating and the tongue body, and 2 clustering clusters are sufficient. However, the tongue images of patients with diabetes nephropathy may have various abnormal manifestations, which can be mainly divided into the tongue body, the tongue proper, the crack and the petechial plaque regions. Select the number of clustering clusters to be 2 to 5, that is, the number of clustering clusters is 2, 3, 4, and 5 respectively. The FCM clustering algorithm is used to cluster the pixel points in the tongue image region respectively, and the clustering result each time and the membership of each pixel point in each cluster in the clustering result each time are obtained. The acquisition processes of the FCM clustering algorithm and the membership are both prior arts and will not be elaborated here. In specific applications, the implementer can set the number of clustering clusters according to the specific situation.

[0056] When performing FCM clustering on the tongue image regions of different patients, the number of clustering clusters should be selected differently according to the manifestations of the tongue image. Too many or too few will lead to improper region segmentation. It should be ensured that the color difference between the pixel points within the class is small, while there is a large color difference between the classes. Therefore, all clustering results are analyzed and the optimal clustering result is selected from them for subsequent analysis and division.

[0057] Specifically, taking the clustering result of one clustering as an example for analysis, the method provided in this embodiment can be used to process other clustering results.

[0058] Record any clustering result as a candidate result, calculate the gray-scale dispersion degree of all pixel points in each cluster in the candidate result, and obtain the gray-scale consistency index of the candidate result according to the gray-scale dispersion degree of all pixel points in all clusters in the candidate result. The gray-scale dispersion degree is negatively correlated with the gray-scale consistency index. Calculate the average gray-scale value of the pixel points in each cluster in the candidate result according to the gray-scale values of all pixel points in each cluster in the candidate result. Obtain the segmentation evaluation value of the candidate result according to the gray-scale consistency index of the candidate result and the dispersion degree of the average gray-scale values of all pixel points in all clusters.

[0059] Among them, in this embodiment, the variance of the gray-scale values of all pixel points in each cluster is used to characterize the gray-scale dispersion degree of all pixel points in each cluster, and the variance of the average gray-scale values of all pixel points in all clusters is used to characterize the dispersion degree of the average gray-scale values of all pixel points in all clusters.

[0060] The negative correlation relationship means that the dependent variable will decrease as the independent variable increases, and the dependent variable will increase as the independent variable decreases. It can be a subtraction relationship, a division relationship, etc., which is determined by the actual application.

[0061] In this embodiment, a specific calculation formula for the segmentation evaluation value is given. The segmentation evaluation value of the clustering result of the b-th clustering can be expressed as:

[0062] ;

[0063] Among them, represents the segmentation evaluation value of the clustering result of the b-th clustering, N represents the number of clusters in the clustering result of the b-th clustering, represents the variance of the gray-scale values of all pixel points in the n-th cluster in the clustering result of the b-th clustering, represents the variance of the average gray-scale values of all pixel points in all clusters in the clustering result of the b-th clustering, represents the exponential function with the natural constant as the base.

[0064] The variance of the gray-scale values of all pixel points in the n-th cluster in the clustering result of the b-th clustering is used to reflect the gray-scale dispersion situation of all pixel points in the n-th cluster. The larger this value is, the more discrete the gray-scale distribution of the pixel points in this cluster is; represents the gray-scale consistency index of the clustering result of the b-th clustering. The larger this value is, the greater the color consistency of the pixel points in the same cluster in the clustering result of the b-th clustering, and the better the segmentation effect; the larger the variance of the average gray-scale values of all pixel points in all clusters in the clustering result of the b-th clustering, the greater the gray-scale difference between different clusters, the better the clustering segmentation effect, and the larger the segmentation evaluation value.

[0065] By using the above method, the segmentation evaluation value of the clustering result of each clustering can be obtained. The larger the segmentation evaluation value is, the more consistent the gray levels of the pixel points within the same cluster are, and the greater the gray level difference between the pixel points in different clusters is, that is, the better the clustering effect. Therefore, the clustering result corresponding to the maximum value of the segmentation evaluation value is taken as the optimal clustering result. As Figure 2 shown, this figure is the optimal clustering result diagram of the tongue image area.

[0066] Step S22: Combine the gray-level gradient rules of the pixel points in different directions within the neighborhood of each pixel point in the tongue image area and the relative position relationship between the pixel point and each cluster in the optimal clustering result to obtain the membership degree of each pixel point in each cluster in the optimal clustering result.

[0067] After screening out the optimal clustering result, there are still problems with inaccurate segmentation of some pixel points in the optimal clustering result. This is because the boundaries between different regions of the tongue image are blurred, and accurate clustering and segmentation of regions cannot be achieved only based on color differences. It is also necessary to adjust the FCM clustering result according to the abnormal manifestation characteristics of the tongue images of diabetic nephropathy patients to accurately segment the region according to the actual tongue image performance. FCM clustering allows each pixel point to belong to multiple clustering clusters with a certain membership degree. By using the tongue image performance characteristics, the membership degree of each pixel point belonging to different clustering clusters is adjusted, so that the pixel points are segmented into the correct regions.

[0068] The abnormal manifestations in each stage of the development of diabetic nephropathy patients include changes in the color of the tongue body, an increase in petechiae or ecchymoses, changes in the color of the tongue coating, an increase in cracks, etc. As diabetic nephropathy develops, the color difference between the tongue coating and the tongue body gradually becomes larger. From the center of the tongue coating to the edge of the tongue, the boundary between the tongue coating and the tongue body is relatively blurred, but the gray-level gradient direction from the center of the tongue coating to the edge of the tongue is consistent. The colors of the pixel points in different regions have a gradient rule, resulting in the color of the pixel points at the region boundary changing to a high similarity with the color of another region, and thus being mis-segmented into other regions. Therefore, when the pixel point conforms to the gray-level gradient rule in the direction from the region center to the pixel point, it indicates that the pixel point is segmented into the correct region.

[0069] Based on this, in this embodiment, the gray-level gradient rules of the pixel points in different directions within the neighborhood of the pixel point will be analyzed to determine the membership degree of each pixel point in each cluster in the optimal clustering result.

[0070] Taking a pixel point in the tongue image area as an example for analysis, the method provided in this embodiment can be used to process other pixel points in the tongue image area.

[0071] Specifically, any pixel point in the tongue image region is denoted as a candidate pixel point, and any direction is denoted as a feature direction. The gray values of the pixel points in the neighborhood of the candidate pixel point in the feature direction and the gray value of the candidate pixel point are respectively subjected to curve fitting to obtain a fitting curve corresponding to the candidate pixel point in the feature direction, and the slope corresponding to each pixel point on the fitting curve is obtained; the abscissa of the fitting curve is the order of the pixel points, and the ordinate is the gray value of the pixel points. The methods for curve fitting and obtaining the slope are both existing technologies and will not be elaborated here.

[0072] According to the slope difference between each pixel point on the fitting curve and its adjacent pixel points, the gray-scale gradient consistency of the candidate pixel point in the feature direction is obtained, and the slope difference is negatively correlated with the gray-scale gradient consistency.

[0073] Among them, the negative correlation means that the dependent variable will decrease as the independent variable increases, and the dependent variable will increase as the independent variable decreases. It can be a subtraction relationship, a division relationship, etc., which is determined by the actual application.

[0074] In this embodiment, a specific calculation formula for the gray-scale gradient consistency is given. The gray-scale gradient consistency of the i-th pixel point in the tongue image region in the f-th direction can be expressed as:

[0075] ;

[0076] Among them, represents the gray-scale gradient consistency of the i-th pixel point in the tongue image region in the f-th direction, represents the total number of the i-th pixel point and the pixel points in its neighborhood in the f-th direction, represents the slope corresponding to the r-th pixel point on the fitting curve corresponding to the i-th pixel point in the f-th direction; represents the slope corresponding to the -th pixel point on the fitting curve corresponding to the i-th pixel point in the f-th direction, represents the absolute value symbol.

[0077] The slope corresponding to the pixel point on the fitting curve is used to reflect the gray-scale change situation, is used to reflect the difference between the slope corresponding to the r-th pixel point and the -th pixel point on the fitting curve corresponding to the i-th pixel point in the f-th direction, represents the average difference of the slopes between adjacent pixel points on the fitting curve corresponding to the i-th pixel point in the f-th direction, which is used to reflect the consistency of the gray-scale change directions of all pixel points in the neighborhood range of the i-th pixel point in the f-th direction. The smaller this value is, the higher the degree of consistency of the gray-scale value change directions in this direction, and the stronger the gray-scale gradient law, that is, the greater the gray-scale gradient consistency.

[0078] By using the above method, the gray-scale gradient consistency of each pixel point in the tongue image region in each direction can be obtained.

[0079] Next, taking the candidate pixel point as an example, the direction from the center point of each cluster in the optimal clustering result to the candidate pixel point is determined as the gray-scale change direction of the candidate pixel point relative to each cluster. Next, according to the gray-scale gradient consistency in the gray-scale change direction of the candidate pixel point relative to each cluster, the membership degree of the candidate pixel point in each cluster in the optimal clustering result is obtained.

[0080] Specifically, any one cluster in the optimal clustering result is denoted as the cluster to be analyzed. According to the difference between the gray-scale gradient consistency of the candidate pixel point in the gray-scale change direction relative to the cluster to be analyzed and the gray-scale gradient consistency of the gray-scale change direction of other clusters except the cluster to be analyzed, the membership degree of the candidate pixel point in the cluster to be analyzed is obtained, and the difference between the gray-scale gradient consistencies is negatively correlated with the membership degree.

[0081] Among them, the negative correlation means that the dependent variable will decrease as the independent variable increases, and the dependent variable will increase as the independent variable decreases. It can be a subtraction relationship, a division relationship, etc., which is determined by the actual application.

[0082] In this embodiment, a specific calculation formula for the membership degree is given. The membership degree of the i-th pixel point in the tongue image region in the j-th cluster in the optimal clustering result can be expressed as:

[0083] ;

[0084] Among them, represents the membership degree of the i-th pixel point in the tongue image region in the j-th cluster in the optimal clustering result, M represents the number of clusters in the optimal clustering result, represents the gray-scale gradient consistency of the i-th pixel point in the tongue image region in the gray-scale change direction relative to the j-th cluster, represents the gray-scale gradient consistency of the i-th pixel point in the tongue image region in the gray-scale change direction of the m-th cluster except the j-th cluster.

[0085] represents the difference between the gray-scale gradient consistency of the i-th pixel point in the gray-scale change direction relative to the j-th cluster and the gray-scale gradient consistency of the i-th pixel point in the gray-scale change direction of the m-th cluster except the j-th cluster, It represents the degree of difference between the gray-scale change direction of the i-th pixel point relative to the j-th cluster and the gray-scale gradual change consistency of the gray-scale change direction of the i-th pixel point in other clusters except the j-th cluster. When this value is larger, it indicates that the gray-scale gradual change rule of the i-th pixel point in the change direction of the j-th cluster is greater than that in other clusters, which means that the membership degree of the i-th pixel point in the j-th cluster in the optimal clustering result is higher.

[0086] By using the above method, the membership degree of each pixel point in the tongue image region in each cluster in the optimal clustering result can be obtained.

[0087] Step S23: According to the difference between the membership degrees of each pixel point and its neighboring pixel points in the same cluster in the optimal clustering result, obtain the membership performance degree of each pixel point and each cluster in the optimal clustering result.

[0088] Considering that there are easily errors or accidental situations in the membership degrees of each pixel point obtained in the optimal clustering result, it is necessary to further confirm the membership performance degree of each pixel point and each cluster according to the similarity degree of the membership degrees of other pixel points in the neighborhood of each pixel point. For any pixel point, when the membership degrees of other pixel points in the neighborhood of this pixel point are more similar to the membership degree of this pixel point to a certain region, it indicates that the membership performance degree of this pixel point to this region is greater.

[0089] Based on this, in this embodiment, the candidate pixel point is still taken as an example for analysis. According to the difference between the membership degree of the candidate pixel point in the cluster to be analyzed and the membership degrees of the pixel points in the neighborhood of the candidate pixel point in the cluster to be analyzed, the membership performance degree of the candidate pixel point and the cluster to be analyzed is obtained, and the difference between the membership degrees is negatively correlated with the membership performance degree.

[0090] Among them, the negative correlation means that the dependent variable will decrease as the independent variable increases, and the dependent variable will increase as the independent variable decreases. It can be a subtraction relationship, a division relationship, etc., which is determined by the actual application.

[0091] In this embodiment, a specific calculation formula for the membership performance degree is given. The membership performance degree of the i-th pixel point in the tongue image region in the j-th cluster in the optimal clustering result can be expressed as:

[0092] ;

[0093] Among them, represents the membership performance degree of the i-th pixel point in the tongue image region in the j-th cluster in the optimal clustering result, represents the membership degree of the i-th pixel point in the tongue image region in the j-th cluster in the optimal clustering result, It represents the mean of the membership degrees of all pixel points in the neighborhood of the i-th pixel point in the tongue image region to the j-th cluster in the optimal clustering result.

[0094] It represents the difference between the membership degree of the i-th pixel point to the j-th cluster in the optimal clustering result and the mean of the membership degrees of other pixel points in its neighborhood to the j-th cluster in the optimal clustering result. The smaller this value is, the more likely the i-th pixel point belongs to the j-th cluster, and the greater its membership performance.

[0095] In this embodiment, for the i-th pixel point, the pixel points in its neighborhood are the pixel points in its eight-neighborhood. In other embodiments, the implementer can set according to specific situations.

[0096] By using the above method, the membership performance of each pixel point to each cluster in the optimal clustering result can be obtained. As Figure 3 shown, this figure is the flowchart of the method for obtaining the membership performance.

[0097] Step S3, according to the membership performance and the initial membership degree of each pixel point to each cluster in the optimal clustering result, the tongue image region is re-divided to obtain different regions.

[0098] Next, in this embodiment, according to the membership performance of each pixel point to each cluster in the optimal clustering result, the membership degree obtained in the FCM clustering is adjusted. The greater the membership performance of each pixel point to each region, the more likely the pixel point belongs to that region. The membership degree of the pixel point in that region is increased so that it can be segmented into that region during the clustering process.

[0099] For any pixel point in the tongue image region: The product of the membership performance of the pixel point to the cluster to be analyzed and the initial membership degree of the pixel point to the cluster to be analyzed is used as the final membership degree of the pixel point to the cluster to be analyzed. By using this method, the final membership degree of each pixel point to each cluster in the optimal clustering result can be obtained. Among them, the initial membership degree is the membership degree of each pixel point to each cluster in the optimal clustering result when obtaining the optimal clustering result. The process of obtaining this membership degree is prior art and will not be elaborated here.

[0100] Next, each pixel point is respectively divided into the cluster with the largest final membership degree corresponding to it to obtain multiple regions, that is, the pixel points in the tongue image region are divided into multiple regions.

[0101] The method provided in this embodiment obtains the adjusted membership degree of each pixel point to each region according to the law of the tongue image manifestation of diabetic nephropathy patients, assigns each pixel point to the clustering cluster region with the largest final membership degree, and obtains the final image segmentation result, improving the accuracy of the segmentation result. Subsequently, doctors can analyze each region separately, improving the analysis efficiency and the accuracy of the evaluation result.

[0102] In this embodiment, first, the tongue image region of the to-be-detected diabetic nephropathy patient is collected, then the number of multiple different clustering clusters is set respectively, and the pixel points in the tongue image region are clustered multiple times. The optimal clustering result is analyzed and screened according to the gray-scale dispersion of the pixel points in each clustering result; for diabetic nephropathy patients, their tongue images will show special manifestations, and the pixel points in the tongue image region will show a certain gray-scale change law. Further, combining the gray-scale gradient law of the pixel points in different directions within the neighborhood of each pixel point in the tongue image region and the relative position relationship between the pixel point and each cluster in the optimal clustering result, the initial membership degree of each pixel point to each cluster in the optimal clustering result is adjusted, so that the pixel points in the tongue image region are divided not only according to color differences, but also combined with the special texture change law of the tongue image of diabetic nephropathy patients, making the division of each pixel point more accurate and the division result of the region more complete and accurate.

[0103] An embodiment of a tongue image auxiliary recognition system for diabetic nephropathy patients:

[0104] As Figure 4 shown, the figure shows the structural block diagram of the tongue image auxiliary recognition system for diabetic nephropathy patients. The system includes an image acquisition module, a membership manifestation calculation module, and a region division module;

[0105] The image acquisition module is used to obtain the tongue image region of the to-be-detected diabetic nephropathy patient;

[0106] The membership manifestation calculation module is used to cluster the pixel points in the tongue image region respectively based on the preset number of different clustering clusters, screen the optimal clustering result according to the gray-scale dispersion of the pixel points in each clustering result; combine the gray-scale gradient law of the pixel points in different directions within the neighborhood of each pixel point in the tongue image region and the relative position relationship between the pixel point and each cluster in the optimal clustering result to obtain the membership degree of each pixel point to each cluster in the optimal clustering result; obtain the membership manifestation degree of each pixel point to each cluster in the optimal clustering result according to the difference between the membership degree of each pixel point and its neighborhood pixel points in the same cluster in the optimal clustering result;

[0107] The region division module is used to re-divide the tongue image region according to the membership manifestation degree and the initial membership degree of each pixel point to each cluster in the optimal clustering result to obtain different regions.

[0108] It should be understood that Figure 4 The structural block diagram and its modules of an auxiliary tongue image recognition system for diabetic nephropathy patients shown can be implemented in various ways. For example, in some embodiments, the system and its modules can be implemented through hardware, software, or a combination of software and hardware. Among them, the hardware part can be implemented using dedicated logic; the software part can be stored in a memory and executed by an appropriate instruction execution system, such as a microprocessor or dedicated design hardware. Those skilled in the art can understand that the above methods and systems can be implemented using computer-executable instructions and / or included in processor control code. For example, such code is provided on a carrier medium such as a disk, CD, or DVD-ROM, a programmable memory such as a read-only memory (firmware), or a data carrier such as an optical or electronic signal carrier. The systems and their modules in this specification can be implemented not only by hardware circuits such as very large scale integrated circuits or gate arrays, semiconductors such as logic chips and transistors, or programmable hardware devices such as field programmable gate arrays and programmable logic devices, but also by software executed by various types of processors, or by a combination of the above hardware circuits and software (for example, firmware).

[0109] More details about the above-mentioned various modules can be referred to other locations in this specification and will not be elaborated here.

[0110] In other embodiments, a medium is also provided, and the medium stores at least one program that can be run by a computer. When the at least one program is run by the computer, the computer is caused to execute the steps in the auxiliary tongue image recognition method for diabetic nephropathy patients in the above embodiments, and the medium can be a computer-readable storage medium.

[0111] Among them, the provided system and medium are both used to execute the corresponding methods provided above. Therefore, the beneficial effects that can be achieved can refer to the beneficial effects in the corresponding methods provided above and will not be elaborated here.

[0112] It should be noted that the above are only the preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent replacements, improvements, etc. made within the principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for assisting in identifying tongue images of patients with diabetic nephropathy, characterized in that: The method comprises the following steps: Acquire a tongue region image of a patient with diabetic nephropathy to be detected; Based on the preset number of different clustering clusters, the pixel points in the tongue image region image are clustered respectively, and the optimal clustering result is selected according to the grayscale discreteness of the pixel points in each clustering result; combining the grayscale gradient law of the pixel points in different directions in the neighborhood of each pixel point in the tongue image region image and the relative position relationship between the pixel point and each cluster in the optimal clustering result, the belonging degree of each pixel point in each cluster in the optimal clustering result is obtained; according to the difference between the belonging degrees of each pixel point and its neighboring pixel points in the same cluster in the optimal clustering result, the belonging expression degree of each pixel point to each cluster in the optimal clustering result is obtained; According to the attribution expression and the initial membership of each pixel point to each cluster in the optimal clustering result, the tongue image region image is divided again to obtain different regions; The degree of belonging of each pixel point to each cluster in the optimal clustering result is obtained, including: Performing curve fitting on the grayscale values ​​of the pixels in the neighborhood of the candidate pixel in the feature direction and the grayscale value of the candidate pixel, respectively, to obtain a fitting curve corresponding to the candidate pixel in the feature direction; According to the slope difference between each pixel point and its adjacent pixel points on the fitting curve, the grayscale gradient consistency of the candidate pixel point in the feature direction is obtained, and the slope difference is negatively correlated with the grayscale gradient consistency; The direction from the center point of each cluster in the optimal clustering result to the candidate pixel point is determined as the grayscale change direction of the candidate pixel point relative to each cluster; According to the grayscale gradient consistency of the candidate pixel points relative to each cluster in the grayscale change direction, the degree of belonging of the candidate pixel points to each cluster in the optimal clustering result is obtained; The candidate pixel point is any pixel point in the tongue image area image, and the characteristic direction is any direction.

2. The method for assisting in identifying tongue images of patients with diabetic nephropathy according to claim 1, characterized in that: The screening of the optimal clustering result according to the grayscale discreteness of the pixel points in each clustering result includes: Calculate the grayscale dispersion degree of all pixels in each cluster of the candidate result, and obtain the grayscale consistency index of the candidate result according to the grayscale dispersion degree of all pixels in all clusters of the candidate result, wherein the grayscale dispersion degree is negatively correlated with the grayscale consistency index; obtain the segmentation evaluation value of the candidate result according to the grayscale consistency index of the candidate result and the dispersion degree of the average grayscale value of the pixels in all clusters; the candidate result is any clustering result; The clustering result corresponding to the maximum value of the segmentation evaluation value is taken as the optimal clustering result.

3. The method for assisting in identifying tongue images of patients with diabetic nephropathy according to claim 1, characterized in that: The step of obtaining the degree of belonging of the candidate pixel to each cluster in the optimal clustering result according to the grayscale gradient consistency of the candidate pixel relative to the grayscale change direction of each cluster includes: The degree of belonging of the candidate pixel to the cluster to be analyzed is obtained according to the difference between the grayscale gradient consistency of the candidate pixel in the grayscale change direction relative to the cluster to be analyzed and the grayscale gradient consistency of the other clusters except the cluster to be analyzed, and the difference between the grayscale gradient consistency and the degree of belonging are negatively correlated; The cluster to be analyzed is any cluster in the optimal clustering result.

4. The method for assisting in identifying tongue images of patients with diabetic nephropathy according to claim 1, characterized in that: The obtaining the belonging expression degree of each pixel point to each cluster in the optimal clustering result according to the difference between the belonging degrees of each pixel point and its neighboring pixel points in the same cluster in the optimal clustering result includes: According to the difference between the degree of attribution of the candidate pixel in the cluster to be analyzed and the degree of attribution of the pixels in the neighborhood of the candidate pixel in the cluster to be analyzed, the attribution expression degree of the candidate pixel and the cluster to be analyzed is obtained, and the difference between the attribution degrees is negatively correlated with the attribution expression degree.

5. The method for assisting in identifying tongue images of patients with diabetic nephropathy according to claim 3, characterized in that: The tongue image is divided again to obtain different regions according to the attribution expression and the initial membership of each pixel point to each cluster in the optimal clustering result, including: For any pixel point in the tongue image area: the product of the attribution expression degree of the pixel point to the cluster to be analyzed and the initial membership degree of the pixel point to the cluster to be analyzed is used as the final membership degree of the pixel point to the cluster to be analyzed; Each pixel point is divided into the cluster with the largest final membership to obtain different areas.

6. The method for assisting in identifying tongue images of patients with diabetic nephropathy according to claim 1, characterized in that: The FCM clustering algorithm is used to cluster all the pixels in the tongue area image.

7. The method for assisting in identifying tongue images of patients with diabetic nephropathy according to claim 1, characterized in that: When the pixels in the tongue image region image are clustered based on the preset number of different clusters, the method further includes: obtaining an initial membership degree between each pixel in the tongue image region image and each cluster in the clustering result.

8. A tongue image auxiliary recognition system for patients with diabetic nephropathy, characterized in that: The system is used to execute the steps in the method for assisting in identifying tongue images of patients with diabetic nephropathy according to any one of claims 1 to 7, and the system comprises: An image acquisition module, used for acquiring tongue image of the diabetic nephropathy patient to be detected; The attribution expression calculation module is used to cluster the pixels in the tongue image region image based on the number of preset different clustering clusters, and select the optimal clustering result according to the grayscale discreteness of the pixels in each clustering result; combine the grayscale gradient of the pixels in different directions in the neighborhood of each pixel in the tongue image region image and the relative position relationship between the pixel and each cluster in the optimal clustering result to obtain the attribution degree of each pixel in each cluster in the optimal clustering result; obtain the attribution expression degree of each pixel and each cluster in the optimal clustering result according to the difference between the attribution degree of each pixel and its neighboring pixels in the same cluster in the optimal clustering result; The region division module is used to divide the tongue image again to obtain different regions according to the attribution expression and the initial membership of each pixel point to each cluster in the optimal clustering result.

9. A medium storing at least one program executable by a computer, characterized in that: When the at least one program is executed by a computer, the computer executes the steps of the method for assisting in identifying tongue images of patients with diabetic nephropathy as described in any one of claims 1 to 7.

Images