Capillary radiation system image processing method

Abstract

The invention relates to the technical field of image processing, in particular to a capillary radiation system image processing method, which comprises the following steps of: acquiring a grey-scale map of a thermal imaging image of a capillary radiation system, and constructing a feature point of each pixel, all the feature points serve as initial class clusters, iteration splitting is carried out through a splitting type hierarchical clustering algorithm, and splitting is carried out according to a splitting point if the preference degree of the splitting point is larger than a preset value during each time of splitting; and if the preset value is smaller than or equal to the preset value, screening the feature points of which the preference degree meets a preset condition from the class cluster where the splitting point is located as new splitting points to perform splitting until a preset termination condition is met, obtaining a splitting result, and performing anomaly detection on the corresponding region according to the anomaly index of each class cluster in the splitting result. According to the invention, the efficiency and accuracy of carrying out anomaly detection on the area where the capillary radiation system is installed can be improved.

Description

Technical Field

[0001] The present invention relates to the field of image processing technology, and more specifically, to a capillary radiation system image processing method. Background Art

[0002] With the rise of green buildings and smart homes, capillary radiation systems have been widely used as efficient and comfortable heating systems. However, when the capillary radiation system is working normally, it uses a water supply pipe to continuously transport flowing hot water to circulate throughout the system. It is usually laid inside the wall. Once problems such as local blockage or structural damage occur, it will cause uneven water temperature distribution, resulting in temperature differences. This temperature difference not only affects the heating effect of the system, but also reduces the user's comfort experience. Therefore, quickly and accurately detecting abnormal areas in the area where the capillary radiation system is installed is of vital importance to ensure the normal operation of the system, improve user experience, and perform timely repairs and maintenance.

[0003] In the related technology, a GIS casing temperature anomaly detection method, device, equipment and readable storage medium are disclosed in a patent application document with publication number CN114049336A. The method includes: obtaining an infrared image marked with a GIS casing area, obtaining the temperatures corresponding to all pixel points in the GIS casing area, and forming a temperature set; constructing a spatial distribution model of the temperature of the GIS casing area, removing invalid data, and smoothing the single-point noise in the infrared image; constructing a frequency distribution model of the temperature of the GIS casing area, setting up a double baseline, dividing the interference area and the normal area, and extracting the main information; marking the pixel points after the noise is removed, and marking the pixel points whose temperatures after the noise is removed exceed the preset threshold as suspected abnormal points, and marking the highest temperature point and the lowest temperature point if the temperature does not exceed the preset threshold.

[0004] However, when marking suspected abnormal points, the above scheme only considers that the higher the temperature, the higher the probability of abnormality in the GIS casing area. But when the pipeline ruptures, the temperature of the corresponding area may be lower than the surrounding temperature. The above scheme will exclude all pixel points with temperatures lower than the low temperature baseline, resulting in missed detection and affecting the accuracy of the detection results. Summary of the invention

[0005] In order to solve the problem of being unable to accurately detect abnormal temperature areas, the present invention provides a capillary radiation system image processing method. The method comprises: Obtain a grayscale image of the thermal imaging image of the capillary radiation system, and construct a feature point of each pixel by using the temperature value of each pixel in the grayscale image as the horizontal coordinate and the pixel position of the corresponding pixel as the vertical coordinate; All feature points are used as initial clusters and iteratively split using a splitting hierarchical clustering algorithm. At each split, if the preference of the split point is greater than a preset value, the split is performed at the split point; if it is less than or equal to the preset value, a feature point whose preference meets the preset conditions is selected from the cluster where the split point is located as a new split point for splitting until the preset termination condition is met to obtain a split result; According to the abnormal index of each cluster in the split result, the corresponding area is detected for abnormality. The abnormal index is the normalized value of the extreme difference of the gradient direction angle of the feature point in the corresponding cluster. The method for obtaining the optimality is as follows: obtaining the degree of change of the temperature value of any feature point, the degree of change represents the difference between the temperature value of the feature point and the mode of the temperature values ​​of all feature points, calculating the credibility of the temperature value of the feature point, the credibility is negatively correlated with the standard deviation of the gradient direction angles of all feature points in the eight neighborhoods, and the sum of the differences in the temperature values ​​of the feature point and each feature point in the eight neighborhoods; The ratio of the change degree to the credibility is taken as the splitting index of the feature point, and the ratio of the splitting index of the feature point to the maximum value of the splitting index of all feature points is calculated to obtain the preference of the feature point.

[0006] When the present invention utilizes a splitting hierarchical clustering algorithm to split the initial clusters composed of all feature points, it is possible to avoid over-segmentation of the initial clusters and improve segmentation efficiency, thereby enabling efficient detection of abnormal temperature areas where a capillary radiation system is installed; and when the initial clusters composed of all feature points are split, it is possible to comprehensively determine the splitting points based on various data, thereby ensuring the accuracy of the splitting results, thereby ensuring the accuracy of the detection results while improving the detection efficiency.

[0007] Preferably, the degree of change satisfies the following relationship: ; In the formula, For the The degree of change of the temperature value of each characteristic point; For the The temperature value of each characteristic point; is the mode of the temperature values ​​of all feature points; is the standard deviation of the temperature values ​​of all characteristic points; is the absolute value symbol.

[0008] The present invention performs normalization through standard deviation, and can eliminate the influence of different unit dimensions.

[0009] Preferably, the credibility of the temperature value of any characteristic point satisfies the following relationship: ; In the formula, For the The credibility of the temperature value of each feature point; For the The first The gradient direction angle of feature points; is the mean of the gradient direction angles of all feature points in the eight neighborhoods; is the number of feature points in the eight neighborhoods; For the The temperature value of each characteristic point; is the first The temperature value of each characteristic point; is a natural exponential function.

[0010] The present invention utilizes the feature that the gradient direction of the temperature values ​​in the normal temperature change area is relatively consistent, while the gradient direction of the temperature values ​​in the abnormal temperature change area is relatively dispersed, and can accurately evaluate the possibility that the temperature value of each feature point is a normal value, so as to eliminate the influence of the normal temperature change area on the abnormal temperature area detection.

[0011] Preferably, at any splitting, if the preference of the splitting point determined by the splitting hierarchical clustering algorithm is less than or equal to a preset value, then it is determined whether the preference of the feature point with the largest splitting index in the cluster where the splitting point is located is greater than the preset value; If it is greater than the preset value, it is determined that the preference of the feature point meets the preset conditions, and the feature point is split as a new splitting point; if it is less than or equal to the preset value, it is determined that the splitting process meets the preset termination conditions, and the splitting process is terminated to obtain the splitting result.

[0012] The method of determining the splitting points of the present invention can avoid the over-segmentation of the initial cluster composed of all feature points, improve the segmentation efficiency, and thus improve the efficiency of anomaly detection.

[0013] Preferably, at any split, the maximum value of the Euclidean distance between any two feature points in the cluster where the split point is located is greater than the maximum value of the Euclidean distance between any two feature points in the remaining clusters.

[0014] Preferably, the extreme difference of the gradient direction angle of the feature points in any cluster is the difference between the maximum value and the minimum value of the gradient direction angle of the feature points in the cluster.

[0015] Preferably, the abnormal index of any cluster in the splitting result satisfies the following relationship: ; In the formula, The split result The anomaly index of each cluster; is the maximum value of the gradient direction angle of the feature points in this cluster; is the minimum value of the gradient direction angle of the feature points in this cluster; is the normalization function.

[0016] Preferably, the method for obtaining the gradient direction angle of any feature point includes: The Sobel operator is used to obtain the gradient direction angle of each feature point.

[0017] Preferably, anomaly detection is performed on the corresponding area according to the anomaly index of each cluster in the split result, including: Obtaining a preset warning threshold, if the abnormal index of any cluster in the splitting result is greater than the abnormal index threshold, it is determined that the capillary network in the area corresponding to the cluster is abnormal, and an alarm is issued; If it is less than or equal to the warning threshold, it is determined that the capillary network in the area corresponding to the cluster is normal.

[0018] The present invention can efficiently and accurately locate the area where abnormalities exist in the capillary network.

[0019] Preferably, at each split, if the preference of the split point and the feature points in the cluster where the split point is located are both less than or equal to a preset value, it is determined that the split process meets the termination condition and the split process is stopped.

[0020] The present invention has the following effects: 1. When the present invention uses a splitting hierarchical clustering algorithm to split the initial clusters composed of all feature points, it can ensure that the preferred degree of the splitting points determined by each splitting operation is the largest and greater than a preset value. Therefore, while being able to divide feature points with similar temperature values ​​into the same cluster, it avoids over-segmentation of clusters composed of feature points in normal temperature change areas, thereby improving the splitting efficiency and thus improving the detection efficiency of abnormal temperature areas in the capillary radiation system.

[0021] 2. The present invention determines the preference of each feature point by integrating various data, thereby ensuring the accuracy of the preference, so that the feature points in the abnormal temperature change area have a greater preference, so that the feature points in the abnormal temperature change area can be segmented out, ensuring the accuracy of the splitting results, thereby improving the accuracy of abnormal detection in the area where the capillary radiation system is installed. BRIEF DESCRIPTION OF THE DRAWINGS

[0022] By reading the following detailed description with reference to the accompanying drawings, the above and other objects, features and advantages of the exemplary embodiments of the present invention will become readily understood. In the accompanying drawings, several embodiments of the present invention are shown in an exemplary and non-limiting manner, and the same or corresponding reference numerals represent the same or corresponding parts, wherein: Figure 1It is a schematic flow chart of the steps of a capillary radiation system image processing method according to an embodiment of the present invention. DETAILED DESCRIPTION

[0023] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative work are within the scope of protection of the present invention.

[0024] The specific embodiments of the present invention are described in detail below with reference to the accompanying drawings.

[0025] Reference Figure 1 , a capillary radiation system image processing method, comprising steps S1 to S3, specifically as follows: S1: Obtain a grayscale image of the thermal imaging image of the capillary radiation system, and construct a feature point of each pixel by using the temperature value of each pixel in the grayscale image as the horizontal coordinate and the pixel position of the corresponding pixel as the vertical coordinate.

[0026] Specifically, a thermal imaging image of the area where the capillary radiation system is installed can be obtained using a thermal imager, and the thermal imaging image can be converted into a grayscale image to simplify the image data for subsequent processing and analysis. It should be noted that the process of converting the thermal imaging image into a grayscale image is a prior art and will not be described in detail in this embodiment.

[0027] Furthermore, the temperature value of each pixel in the grayscale image can be used as the horizontal coordinate, and the position information of the corresponding pixel in the grayscale image can be used as the vertical coordinate, so as to construct the feature point of each pixel in the grayscale image to provide a data basis for subsequent analysis.

[0028] It should be noted that the present invention constructs feature points based on the temperature value and pixel position of each pixel in the grayscale image, and can use the temperature information of all pixels to identify abnormal temperature areas, and accurately locate the identified abnormal temperature areas by pixel positions, thereby effectively detecting abnormalities in the area where the capillary radiation system is installed.

[0029] S2: All feature points are used as initial clusters for iterative splitting using a divisive hierarchical clustering algorithm. During each split, if the preference of the split point is greater than a preset value, the split is performed at the split point; if it is less than or equal to the preset value, feature points whose preference meets the preset conditions are selected from the cluster where the split point is located as new split points for splitting until the preset termination conditions are met to obtain the split result.

[0030] It should be noted that since the temperature characteristics of the area where the capillary radiation system is installed are more obvious, when there is an abnormality in the capillary network, such as blockage or rupture, it usually leads to uneven temperature distribution in the local area, making the temperature distribution in the abnormal area significantly different from that in other normal areas. The divisive hierarchical clustering algorithm can effectively identify areas with uneven temperature distribution by gradually splitting clusters, so that the divisive hierarchical clustering algorithm can be used to realize abnormality detection of the capillary network.

[0031] It should be further explained that the process of selecting split points in the traditional split hierarchical clustering algorithm is: select the object with the largest average dissimilarity from the cluster with the largest diameter as the split point. The diameter of the cluster is the maximum value of the Euclidean distance between any two objects in the corresponding cluster; the average dissimilarity is the average distance between any object in the corresponding cluster and the remaining objects.

[0032] However, when the capillary network is working normally, there may be normal temperature changes. For example, the opening and closing operations of windows, doors and other openings will reduce the temperature of the corresponding area, so that the temperature of the corresponding area is significantly different from that of other areas. The traditional splitting hierarchical clustering algorithm may divide the area with normal temperature changes into multiple clusters, which will lead to the occurrence of over-segmentation and reduce the detection efficiency. Therefore, the present invention improves the algorithm, and the specific improvements are as follows: based on the temperature characteristics of the feature points, and the distinguishing features of normal temperature changes and abnormal temperature changes, the preference of the feature points is calculated, and the splitting points determined by the traditional splitting hierarchical clustering algorithm are judged by the preference. If the preference of the splitting point meets the conditions, the splitting point is split. Otherwise, the feature point with the preference that meets the conditions is re-selected from the cluster where the splitting point is located as a new splitting point, and the splitting is performed based on the new splitting point.

[0033] Specifically, the preference of any feature point can be determined by the following steps: Step 1: Obtain the degree of change of the temperature value of any feature point, where the degree of change represents the difference between the temperature value of the feature point and the mode of the temperature values ​​of all feature points; It should be noted that when the capillary radiation system works normally, the water supply pipe is used to transport hot water to circulate throughout the capillary network, and the temperature distribution is relatively uniform. However, when the capillary network is blocked or ruptured, it will cause obvious heat accumulation or heat dispersion in the local area, thereby making the water temperature distribution uneven. Therefore, the present invention uses the mode of the temperature value as a measurement indicator, so as to analyze the change of the temperature value of each characteristic point compared with the normal temperature value.

[0034] In an exemplary embodiment of the present invention, the difference between the temperature value of any characteristic point and the mode of the temperature values ​​of all characteristic points is a comprehensive difference, and the degree of change of any characteristic point can be determined by the following steps. Specifically, the degree of change of any characteristic point satisfies the following relationship: ; In the formula, For the The degree of change of the temperature value of each characteristic point; For the The temperature value of each characteristic point; is the mode of the temperature values ​​of all feature points; is the standard deviation of the temperature values ​​of all characteristic points, which is used to eliminate the influence of unit dimension; is the absolute value symbol.

[0035] in, Reflects the The deviation of the temperature value of the characteristic point from the normal temperature value. The larger the value, the The capillary network at the position corresponding to the feature point is more likely to be blocked or broken, causing the temperature at this position to deviate significantly from the normal temperature level, and the corresponding degree of change of this feature point is relatively large.

[0036] Step 2: Calculate the credibility of the temperature value of the feature point. The credibility is negatively correlated with the standard deviation of the gradient direction angles of all feature points in the eight neighborhoods and the sum of the temperature differences between the feature point and each feature point in the eight neighborhoods. It should be noted that when the capillary network is working normally, the temperature changes caused by the switch operation of the opening parts such as windows and doors usually have a relatively consistent directionality; when the pipe network is blocked, broken, etc., the direction of the temperature change is usually divergent from the high temperature area to the low temperature area, showing a radial distribution centered on the high temperature area. Therefore, the present invention uses this feature to calculate the credibility of the temperature value of the feature point, which can accurately measure the possibility that the temperature value of each feature point is a normal value, thereby effectively eliminating the influence of the corresponding feature points in the normal temperature change area on the abnormal detection results.

[0037] Specifically, the credibility of the temperature value of any characteristic point satisfies the following relationship: ; In the formula, For the The credibility of the temperature value of each feature point; For the The first The gradient direction angle of feature points; is the mean of the gradient direction angles of all feature points in the eight neighborhoods; is the number of feature points in the eight neighborhoods; For the The temperature value of each characteristic point; is the first The temperature value of each characteristic point; is a natural exponential function.

[0038] in, Reflects the The standard deviation of the gradient direction angles of all feature points in the eight neighborhoods of a feature point. The smaller the value, the more consistent the gradient direction of the temperature value corresponding to the feature point in the eight neighborhoods of the feature point, which means that the feature point is more likely to be in the normal area, and the credibility of the temperature value of the feature point is relatively large.

[0039] Reflects the The sum of the temperature differences between the feature point and the feature points in the eight neighborhoods. The smaller the value, the more consistent the temperature values ​​in the neighborhood of the feature point, which means that the feature point is more likely to be in the normal area, and the credibility of the temperature value of the feature point is relatively large.

[0040] Optionally, the absolute value of the difference between the temperature values ​​of any feature point and each feature point in the eight neighborhoods may be used as the difference between the temperature values ​​of the feature point and the corresponding feature points in the eight neighborhoods. This embodiment does not specifically limit the method for determining the difference.

[0041] In another embodiment, the credibility of the temperature value of any characteristic point may also be calculated by other calculation formulas, such as: ; In the formula, For the The credibility of the temperature value of each feature point; For the The standard deviation of the gradient direction angles of all feature points in the eight neighborhoods of a feature point; For the The temperature value of each characteristic point; is the first The temperature value of each characteristic point; is the number of feature points in the eight neighborhoods; is the absolute value symbol.

[0042] In an exemplary embodiment of the present invention, the determination of the gradient direction angle of any feature point can be achieved by the following steps: The Sobel operator is used to obtain the gradient direction angle of each feature point.

[0043] It should be noted that the process of obtaining the gradient direction angle using the Sobel operator is a prior art and will not be described in detail in this embodiment.

[0044] Optionally, other operators, such as the Prewitt operator, may be used to calculate the gradient direction angle of each feature point.

[0045] Step 3: Take the ratio of the change degree to the credibility as the splitting index of the feature point, calculate the ratio of the splitting index of the feature point to the maximum value of the splitting index of all feature points, and obtain the preference of the feature point.

[0046] The splitting index refers to the possibility of any feature point being a splitting point. It should be noted that when the credibility of the temperature value of any feature point is small and the degree of change of the temperature value of the feature point is large, it means that the temperature value of the feature point is likely to be abnormal, and the corresponding possibility of the feature point being a splitting point is relatively large, that is, the splitting index of the feature point is relatively large. Therefore, the present invention uses the ratio of the degree of change of the temperature value of any feature point to the credibility of the temperature value of the feature point as the splitting index of the feature point, so as to accurately measure the possibility of each feature point being a splitting point.

[0047] It should be further explained that when the splitting index of any feature point is closer to the maximum value of the splitting index of all feature points, the feature point is more likely to be a splitting point, and the corresponding preference of the feature point is greater. Therefore, the present invention uses the ratio of the splitting index of any feature point to the maximum value of the splitting index of all feature points as the preference of the feature point, so that the splitting point can be determined based on the preference each time the splitting operation is performed, avoiding over-segmentation, improving segmentation efficiency, and thus improving the efficiency of anomaly detection.

[0048] Next, the process of iteratively splitting the initial clusters composed of all feature points using the improved split hierarchical clustering algorithm is described in detail: Step 1: In the first splitting process, the traditional splitting hierarchical clustering algorithm is used to determine the splitting point in this splitting process.

[0049] In an exemplary embodiment of the present invention, at any split, the maximum value of the Euclidean distance between any two feature points in the cluster where the split point is located is greater than the maximum value of the Euclidean distance between any two feature points in the remaining clusters. That is, the diameter of the cluster where the split point is located is the largest, which further indicates that the probability that the cluster where the split point is located needs to be split is higher than that of the remaining clusters.

[0050] It should be noted that the process of determining the splitting points using the traditional divisive hierarchical clustering algorithm is a prior art and will not be described in detail in this embodiment.

[0051] Step 2: Determine whether the preference of the splitting point is greater than a preset value. If it is greater than the preset value, perform a splitting operation based on the splitting point to obtain two clusters; if it is less than or equal to the preset value, select feature points from the initial cluster whose preference meets the preset conditions as new splitting points for splitting operations to obtain two clusters; and repeat the splitting process until the preset termination condition is met to obtain a splitting result. In this embodiment, the preset value is 0.6. Of course, a preset value of a suitable size can also be selected according to the specific situation. This embodiment does not specifically limit the value of the preset value.

[0052] In an exemplary embodiment of the present invention, at any splitting, if the preference of the splitting point determined by the divisive hierarchical clustering algorithm is less than or equal to a preset value, it is determined whether the preference of the feature point with the largest splitting index in the cluster where the splitting point is located is greater than the preset value; if it is greater than the preset value, it is determined that the preference of the feature point meets the preset conditions, and the feature point is split as a new splitting point; if it is less than or equal to the preset value, it is determined that the splitting process meets the preset termination conditions, and the splitting process is terminated to obtain the splitting result.

[0053] It should be noted that if there is no feature point with a preference greater than a preset value in the cluster where the splitting point is located, it means that the temperature values ​​of all feature points in the cluster where the splitting point is located are likely to be temperature values ​​in the normal area, and there is no need to perform the splitting operation again, which avoids over-segmentation and improves the splitting efficiency, thereby improving the efficiency of anomaly detection in the corresponding areas of each cluster in the splitting result.

[0054] S3: Perform anomaly detection on the corresponding area according to the anomaly index of each cluster in the split result. The anomaly index is the normalized value of the extreme difference of the gradient direction angle of the feature point in the corresponding cluster.

[0055] It should be noted that when the pipe network is blocked, the water flow slows down, heat accumulates to form a local high temperature zone, and the temperature gradient direction diverges from the high temperature zone to the surrounding low temperature zone, forming a radial distribution; when the pipe network ruptures and causes water leakage, heat is dissipated to form a local low temperature zone, and the temperature gradient direction converges from the surrounding high temperature zone to the low temperature zone, forming a convergent distribution; the temperature gradient direction of the abnormal area usually shows a dispersed or convergent feature, while the temperature gradient direction of the normal area is often relatively consistent. Therefore, the present invention uses this feature to calculate the abnormal index of each cluster in the splitting result, which can accurately evaluate the abnormal conditions of the corresponding areas of each cluster, thereby accurately locating the abnormal area.

[0056] Specifically, the anomaly index of any cluster in the split result satisfies the following relationship: ; In the formula, The split result The anomaly index of each cluster; is the maximum value of the gradient direction angle of the feature points in this cluster; is the minimum value of the gradient direction angle of the feature points in this cluster; is the normalization function.

[0057] in, Reflects the split result The maximum value of the gradient direction angle of the feature points in a cluster is larger, indicating that the gradient direction angle of the temperature value of the feature points in the cluster is more dispersed, which further indicates that the possibility that the corresponding area of ​​the cluster is an abnormal area is relatively large, and the corresponding abnormal index of the cluster is larger.

[0058] In another embodiment, the value range of the hyperbolic tangent function may be limited to 0-1, so as to use the hyperbolic tangent function in the value range of 0-1 to normalize the extreme difference of the gradient direction angle of the feature points in each cluster.

[0059] In an exemplary embodiment of the present invention, the detection of abnormal areas can be achieved by the following steps: Get the preset warning threshold. If the abnormal index of any cluster in the splitting result is greater than the abnormal index threshold, it is determined that the capillary network in the area corresponding to the cluster is abnormal and an alarm is issued; if it is less than or equal to the warning threshold, it is determined that the capillary network in the area corresponding to the cluster is normal.

[0060] Optionally, the warning threshold can be set to 0.8. When the abnormal index of any cluster in the splitting result is less than or equal to 0.8, the capillary network at the corresponding position of all feature points in the cluster is judged to be normal; and when the abnormal index of any cluster in the splitting result exceeds 0.8, the capillary network at the corresponding position of all feature points in the cluster is judged to have defects, such as blockage or rupture, and an alarm is triggered. Then, the abnormal area can be located by the position of each feature point in the cluster, so as to facilitate subsequent maintenance work and realize abnormal detection of the area where the capillary radiation system is installed. This embodiment does not specifically limit the size of the warning threshold.

[0061] In the description of this specification, "plurality" or "several" means at least two, such as two, three or more, etc., unless otherwise clearly and specifically defined.

[0062] Although this specification has shown and described a number of embodiments of the present invention, it will be apparent to those skilled in the art that such embodiments are provided by way of example only. Those skilled in the art will conceive of many modifications, changes and alternatives without departing from the ideas and spirit of the present invention. It should be understood that in the practice of the present invention, various alternatives to the embodiments of the present invention described herein may be employed.

Claims

1. A capillary radiation system image processing method, characterized in that: include: Obtain a grayscale image of the thermal imaging image of the capillary radiation system, and construct a feature point of each pixel by using the temperature value of each pixel in the grayscale image as the horizontal coordinate and the pixel position of the corresponding pixel as the vertical coordinate; All feature points are used as initial clusters and iteratively split using a splitting hierarchical clustering algorithm. At each split, if the preference of the split point is greater than a preset value, the split is performed at the split point; if it is less than or equal to the preset value, a feature point whose preference meets the preset conditions is selected from the cluster where the split point is located as a new split point for splitting until the preset termination condition is met to obtain a split result; According to the abnormal index of each cluster in the split result, the corresponding area is detected for abnormality. The abnormal index is the normalized value of the extreme difference of the gradient direction angle of the feature point in the corresponding cluster. The method for obtaining the preferred degree is as follows: obtaining the degree of change of the temperature value of any feature point, the degree of change represents the difference between the temperature value of the feature point and the mode of the temperature values ​​of all feature points, calculating the credibility of the temperature value of the feature point, the credibility is negatively correlated with the standard deviation of the gradient direction angles of all feature points in the eight neighborhoods, and the sum of the differences in the temperature values ​​of the feature point and each feature point in the eight neighborhoods; The ratio of the change degree to the credibility is taken as the splitting index of the feature point, and the ratio of the splitting index of the feature point to the maximum value of the splitting index of all feature points is calculated to obtain the preference of the feature point.

2. The capillary radiation system image processing method according to claim 1, characterized in that: The degree of change satisfies the following relationship: ； In the formula, For the The degree of change of the temperature value of each characteristic point; For the The temperature value of each feature point; is the mode of the temperature values ​​of all feature points; is the standard deviation of the temperature values ​​of all characteristic points; is the absolute value symbol.

3. The capillary radiation system image processing method according to claim 1, characterized in that: The credibility of the temperature value of any characteristic point satisfies the following relationship: ； In the formula, For the The credibility of the temperature value of each feature point; For the The first The gradient direction angle of feature points; is the mean of the gradient direction angles of all feature points in the eight neighborhoods; is the number of feature points in the eight neighborhoods; For the The temperature value of each feature point; The first The temperature value of each feature point; is a natural exponential function.

4. The capillary radiation system image processing method according to claim 1, characterized in that: At any split, if the preference of the split point determined by the splitting hierarchical clustering algorithm is less than or equal to the preset value, then determine whether the preference of the feature point with the largest split index in the cluster where the split point is located is greater than the preset value; If it is greater than the preset value, it is determined that the preference of the feature point meets the preset conditions, and the feature point is split as a new splitting point; if it is less than or equal to the preset value, it is determined that the splitting process meets the preset termination conditions, and the splitting process is terminated to obtain the splitting result.

5. The capillary radiation system image processing method according to claim 4, characterized in that: At any split, the maximum value of the Euclidean distance between any two feature points in the cluster where the split point is located is greater than the maximum value of the Euclidean distance between any two feature points in the remaining clusters.

6. The capillary radiation system image processing method according to claim 1, characterized in that: The extreme value of the gradient direction angle of the feature points in any cluster is the difference between the maximum value and the minimum value of the gradient direction angle of the feature points in the cluster.

7. The capillary radiation system image processing method according to claim 1, characterized in that: The anomaly index of any cluster in the split result satisfies the following relationship: ； In the formula, The split result is The anomaly index of each cluster; is the maximum value of the gradient direction angle of the feature points in this cluster; is the minimum value of the gradient direction angle of the feature points in this cluster; is the normalization function.

8. The capillary radiation system image processing method according to claim 7, characterized in that: The method for obtaining the gradient direction angle of any feature point includes: The Sobel operator is used to obtain the gradient direction angle of each feature point.

9. The capillary radiation system image processing method according to claim 1, characterized in that: The abnormality detection of the corresponding area according to the abnormality index of each cluster in the split result includes: Obtaining a preset warning threshold, if the abnormality index of any cluster in the splitting result is greater than the abnormality index threshold, it is determined that there is an abnormality in the capillary network in the area corresponding to the cluster, and an alarm is issued; If it is less than or equal to the warning threshold, it is determined that the capillary network in the area corresponding to the cluster is normal.

10. The capillary radiation system image processing method according to claim 1, characterized in that: At each split, if the preference of the split point and the feature points in the cluster where the split point is located is less than or equal to the preset value, it is determined that the split process meets the termination condition and the split process is stopped.

Images