Photovoltaic panel defect inspection method and system based on unmanned aerial vehicle

Abstract

The invention relates to the technical field of photovoltaic power stations, and discloses a photovoltaic panel defect inspection method and system based on an unmanned aerial vehicle, and the method comprises the steps: carrying out the graying processing of the image data of a photovoltaic array, and obtaining a photovoltaic array gray image; performing average segmentation on the photovoltaic array grayscale image to obtain a plurality of sub-array images, and performing image enhancement on the photovoltaic array grayscale image according to the sub-array images; clustering the pixel points according to the gray values of the pixel points of the photovoltaic array gray image after image enhancement; extracting an abnormal sub-array image according to the clustering cluster of the pixel points, performing edge detection on the photovoltaic panel in the abnormal sub-array image, and extracting an abnormal edge region of the photovoltaic panel according to an edge detection result; and calculating a defect characteristic value of the abnormal edge region, and determining a defect type of the photovoltaic panel according to the defect characteristic value of the abnormal edge region. The accuracy of photovoltaic panel defect inspection can be effectively improved, and the inspection efficiency is improved.

Description

Technical Field

[0001] The present application relates to the technical field of photovoltaic power stations, and more specifically, to a method and system for inspecting photovoltaic panel defects based on drones. Background Art

[0002] In order to ensure the safety and efficient power generation of photovoltaic power stations, inspection has become an indispensable task. For photovoltaic power generation, photovoltaic panel defects such as cracks, pollutants or shadows will affect the power generation revenue by 5% to 30%. Therefore, photovoltaic panel inspection is a key factor affecting the investment and revenue of photovoltaic power stations. With the continuous development of science and technology, drone technology has been widely used in various fields. In the energy field, especially in the operation and maintenance management of photovoltaic power stations, drone inspection has become a new type of inspection method that is efficient, safe and environmentally friendly.

[0003] In the prior art, when drones are used to collect images of photovoltaic panels in photovoltaic power stations for inspection, in order to prevent missed detections or false detections, each photovoltaic panel needs to be analyzed one by one, and the collected photovoltaic panel images are easily affected by light, resulting in low accuracy and efficiency of defect detection. Summary of the invention

[0004] The present invention provides a photovoltaic panel defect inspection method and system based on an unmanned aerial vehicle, which is used to solve the problem of low accuracy and efficiency of photovoltaic panel defect detection in the prior art, including:

[0005] Acquire image data of the photovoltaic array, perform grayscale processing on the image data of the photovoltaic array, and obtain a grayscale image of the photovoltaic array;

[0006] The photovoltaic array grayscale image is evenly divided to obtain a number of sub-array images, and the photovoltaic array grayscale image is enhanced according to the sub-array images;

[0007] Obtaining grayscale values ​​of pixel points of the photovoltaic array grayscale image after image enhancement, and clustering the pixel points according to the grayscale values ​​of the pixel points of the photovoltaic array grayscale image after image enhancement;

[0008] Extracting the sub-array image with abnormalities according to the clustering clusters of pixel points, performing edge detection on the photovoltaic panel in the sub-array image with abnormalities, and extracting the abnormal edge area of ​​the photovoltaic panel according to the edge detection result;

[0009] The defect characteristic value of the abnormal edge area is calculated, and the defect type of the photovoltaic panel is determined according to the defect characteristic value of the abnormal edge area.

[0010] Further, the performing image enhancement on the photovoltaic array grayscale image according to the sub-array image comprises:

[0011] Obtaining the grayscale value of the pixel in the subarray image, calculating the average grayscale value of the subarray image, and determining the difference parameter of the pixel according to the difference between the grayscale value of the pixel in the subarray image and the average grayscale value;

[0012] Obtain solar radiation data corresponding to the photovoltaic array grayscale image, and determine the shadow correction coefficient of the pixel point at each position in the photovoltaic array grayscale image according to the solar radiation data;

[0013] The difference parameter of the pixel point is multiplied by the shadow correction coefficient to obtain the corrected difference parameter, and the photovoltaic array grayscale image is enhanced according to the corrected difference parameter.

[0014] Furthermore, the step of determining the shadow correction coefficient of the pixel at each position in the photovoltaic array grayscale image according to the solar radiation data includes:

[0015] Determine the solar irradiation angle at each position of the photovoltaic array according to the solar irradiation data, set the initial irradiation area according to the solar irradiation angle, and determine the distance parameter value of the pixel point according to the distance between the pixel point in the photovoltaic array and the initial irradiation area;

[0016] Determine the light intensity value according to the solar radiation data, and set the angle parameter value according to the light intensity value;

[0017] The angle parameter value is used as the hypotenuse angle, the distance parameter value of the pixel point is converted into the hypotenuse length according to the preset ratio, and a right triangle is constructed according to the hypotenuse angle and the maximum hypotenuse length;

[0018] Take the non-right-angled point of the right triangle close to the base as the initial point, determine the position of each pixel on the hypotenuse of the right triangle according to the length of the hypotenuse of each pixel, determine the height value of each pixel according to the position of each pixel on the hypotenuse of the right triangle, and determine the shadow correction coefficient according to the height value of each pixel.

[0019] Furthermore, the step of performing image enhancement on the photovoltaic array grayscale image according to the corrected difference parameter includes:

[0020] Based on the image enhancement formula, the grayscale value of the pixel point of the photovoltaic array grayscale image is enhanced or suppressed according to the corrected difference parameter. The image enhancement formula is specifically:

[0021]

[0022] Among them, G is the gray value of the pixel after enhancement or suppression, G α is the initial gray value of the pixel, D is the difference parameter after pixel correction, D α is the preset allowable difference parameter, R is the preset constant, and exp is the natural exponential function.

[0023] Furthermore, clustering the pixels according to the grayscale values ​​of the pixels of the photovoltaic array grayscale image after image enhancement includes:

[0024] A grayscale data set is established according to the grayscale values ​​of the pixels of the photovoltaic array grayscale image after image enhancement, and k initial clustering centers of the grayscale data set are randomly selected;

[0025] Calculate the Euclidean distance from the grayscale value in the grayscale data set to the initial cluster center, and divide each pixel point into the corresponding cluster cluster according to the Euclidean distance from the grayscale value in the grayscale data set to the initial cluster center;

[0026] Calculate the grayscale average value in each cluster, and re-determine the cluster center based on the grayscale average value in each cluster;

[0027] Repeat the above steps until the cluster center no longer changes or the number of iterations reaches the preset iteration threshold, and obtain k final cluster centers. According to the clustering results, determine the cluster to which each pixel belongs.

[0028] Furthermore, extracting the sub-array image with abnormalities according to the clustering of pixel points includes:

[0029] Obtain a preset grayscale anomaly threshold, and calculate the difference between the grayscale value of the cluster center of each cluster and the preset grayscale anomaly threshold;

[0030] If the difference between the gray value of the cluster center and the preset gray abnormality threshold is less than the first preset threshold, it is determined that the subarray image corresponding to the cluster cluster has an abnormality;

[0031] If the difference between the grayscale value of the cluster center and the preset grayscale abnormality threshold is greater than or equal to the first preset threshold, the difference between the average grayscale value of the subarray image and the average grayscale value of all adjacent subarray images is determined; if the difference between the average grayscale value of the subarray image and the average grayscale value of all adjacent subarray images is less than the second preset threshold, it is determined that the corresponding subarray image has an abnormality.

[0032] Furthermore, the performing edge detection on the photovoltaic panel in the sub-array image with abnormality and extracting the abnormal edge region of the photovoltaic panel according to the edge detection result includes:

[0033] Based on the canny edge detection algorithm, the edge of the photovoltaic panel in the sub-array image with abnormality is detected to obtain the edge line of the photovoltaic panel;

[0034] Obtaining a preset standard edge line, and calculating the matching degree between the preset standard edge line and the edge line of the photovoltaic panel;

[0035] The edge area where the matching degree between the preset standard edge line and the edge line of the photovoltaic panel is less than the third preset threshold is screened out to obtain the abnormal edge area of ​​the photovoltaic panel.

[0036] Furthermore, the calculation of the defect characteristic value of the abnormal edge area includes:

[0037] Obtaining an average grayscale value of an abnormal edge region of the photovoltaic panel, calculating a difference between an average grayscale value of the remaining region of the photovoltaic panel and an average grayscale value of the abnormal edge region, and obtaining a first abnormal parameter;

[0038] The inverse of the matching degree between the edge line of the abnormal edge area and the standard edge line is set as the second abnormal parameter, and the product of the first abnormal parameter and the second abnormal parameter is calculated to obtain the defect characteristic value of the abnormal edge area.

[0039] Furthermore, the step of determining the defect type of the photovoltaic panel according to the defect characteristic value of the abnormal edge area includes:

[0040] Obtaining a preset allowable characteristic value, and calculating the difference between the defect characteristic value of the abnormal edge area and the preset allowable characteristic value;

[0041] If the difference between the defect characteristic value of the abnormal edge area and the preset allowable characteristic value is greater than a fourth preset threshold, it is determined that the defect type of the photovoltaic panel corresponding to the abnormal edge area is a crack defect;

[0042] If the difference between the defect characteristic value of the abnormal edge area and the preset allowable characteristic value is less than or equal to the fourth preset threshold, it is determined that the photovoltaic panel defect type corresponding to the abnormal edge area is an occlusion defect.

[0043] In order to achieve the above object, the present invention also provides a photovoltaic panel defect inspection system based on a drone, comprising:

[0044] The first module is used to obtain the image data of the photovoltaic array, perform grayscale processing on the image data of the photovoltaic array, and obtain a grayscale image of the photovoltaic array;

[0045] The second module is used to evenly divide the photovoltaic array grayscale image to obtain a number of sub-array images, and perform image enhancement on the photovoltaic array grayscale image according to the sub-array images;

[0046] The third module is used to obtain the grayscale values ​​of the pixels of the photovoltaic array grayscale image after the image enhancement, and cluster the pixels according to the grayscale values ​​of the pixels of the photovoltaic array grayscale image after the image enhancement;

[0047] The fourth module is used to extract the sub-array image with abnormalities according to the clustering clusters of pixel points, perform edge detection on the photovoltaic panel in the sub-array image with abnormalities, and extract the abnormal edge area of ​​the photovoltaic panel according to the edge detection result;

[0048] The fifth module is used to calculate the defect characteristic value of the abnormal edge area and determine the defect type of the photovoltaic panel according to the defect characteristic value of the abnormal edge area.

[0049] The beneficial effects of the present invention are:

[0050] By applying the above technical scheme, the present invention enhances the grayscale image of the photovoltaic array and improves the grayscale contrast of the grayscale image of the photovoltaic array, which can effectively increase the accuracy of abnormality recognition in the photovoltaic array, extract the abnormal image of the photovoltaic array through a clustering algorithm, and extract the abnormal edges of the photovoltaic panel from the abnormal image. There is no need to detect each photovoltaic panel one by one, which improves the inspection efficiency, and then judges the defect type of the photovoltaic panel through the abnormal edge, thereby meeting the requirements of efficient and intelligent inspection. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] In order to more clearly illustrate the technical solutions in the embodiments of the present application, the drawings required for use in the description of the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For those skilled in the art, other drawings can be obtained based on these drawings without creative work.

[0052] Figure 1 A schematic diagram of a process of a photovoltaic panel defect inspection method based on a drone proposed in an embodiment of the present invention is shown;

[0053] Figure 2 The overall structure diagram of a photovoltaic panel defect inspection system based on a drone proposed in an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0054] The following will be combined with the drawings in the embodiments of the present application to clearly and completely describe the technical solutions in the embodiments of the present application. Obviously, the described embodiments are only part of the embodiments of the present application, not all of the embodiments. Based on the embodiments in the present application, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of this application.

[0055] The present application embodiment provides a photovoltaic panel defect inspection method based on a drone, such as Figure 1 As shown, including:

[0056] S101, acquiring image data of a photovoltaic array, performing grayscale processing on the image data of the photovoltaic array, and obtaining a grayscale image of the photovoltaic array;

[0057] In this embodiment, the inspection route of the drone is preset, and the image data of the photovoltaic array is collected based on the high-definition camera of the drone.

[0058] S102, performing average segmentation on the photovoltaic array grayscale image to obtain a plurality of sub-array images, and performing image enhancement on the photovoltaic array grayscale image according to the sub-array images;

[0059] In some embodiments of the present application, the image enhancement of the photovoltaic array grayscale image based on the sub-array image includes: obtaining the grayscale value of the pixel in the sub-array image, calculating the average grayscale value of the sub-array image, and determining the difference parameter of the pixel according to the difference between the grayscale value of the pixel in the sub-array image and the average grayscale value; obtaining solar radiation data corresponding to the photovoltaic array grayscale image, and determining the shadow correction coefficient of the pixel at each position in the photovoltaic array grayscale image according to the solar radiation data; multiplying the difference parameter of the pixel by the shadow correction coefficient to obtain the corrected difference parameter, and performing image enhancement on the photovoltaic array grayscale image according to the corrected difference parameter.

[0060] In this embodiment, after the photovoltaic array image data is gray-scaled, the photovoltaic array gray-scale image is divided into a number of N×N sub-array images by average segmentation, the difference parameter of the pixel is calculated by the difference between the gray-scale value of each pixel in the sub-array image and the average gray-scale value, the difference parameter is corrected by the solar radiation data during the photovoltaic array image acquisition, and the photovoltaic array gray-scale image is enhanced by the corrected difference parameter, thereby achieving image enhancement of the photovoltaic array gray-scale image.

[0061] In some embodiments of the present application, the method of determining the shadow correction coefficient of the pixel points at each position in the grayscale image of the photovoltaic array according to the solar illumination data includes: determining the solar illumination angle of each position of the photovoltaic array according to the solar illumination data, setting the initial illumination area according to the solar illumination angle, and determining the distance parameter value of the pixel point according to the distance between the pixel point in the photovoltaic array and the initial illumination area; determining the light intensity value according to the solar illumination data, and setting the angle parameter value according to the light intensity value; using the angle parameter value as the hypotenuse angle, converting the distance parameter value of the pixel point into the hypotenuse length according to a preset ratio, and constructing a right triangle according to the hypotenuse angle and the maximum hypotenuse length; using the non-right angle point of the right triangle close to the base as the initial point, determining the position of each pixel point on the hypotenuse of the right triangle according to the hypotenuse length of each pixel point, determining the height value of each pixel point according to the position of each pixel point on the hypotenuse of the right triangle, and determining the shadow correction coefficient according to the height value of each pixel point.

[0062] In this embodiment, the optimal solar irradiation position of the photovoltaic array is determined by the solar irradiation angle, the photovoltaic array area at the optimal solar irradiation position is set as the initial illumination area, the distance parameter value is calculated by the distance between the photovoltaic arrays at other positions and the initial illumination area, the angle parameter value is set by the light intensity-angle parameter value mapping table, the greater the light intensity, the greater the corresponding angle parameter value, the range of the angle parameter value is set to (0°~60°), the angle between the hypotenuse and the base is set according to the angle parameter value, the hypotenuse length is set according to the hypotenuse length converted from the maximum distance parameter value to construct a right triangle, each pixel point is assigned to the position corresponding to the hypotenuse according to the hypotenuse length converted from each pixel point, the height value of the pixel point in the right triangle is normalized, the normalized height value is set as the shadow correction coefficient, the difference parameter is corrected by the shadow correction coefficient, and the influence of different brightness caused by different illumination angles on the photovoltaic array is corrected.

[0063] In some embodiments of the present application, the step of performing image enhancement on the photovoltaic array grayscale image according to the corrected difference parameter includes: performing grayscale value enhancement or suppression on the pixel points of the photovoltaic array grayscale image according to the corrected difference parameter based on an image enhancement formula, wherein the image enhancement formula is specifically:

[0064]

[0065] Among them, G is the gray value of the pixel after enhancement or suppression, G α is the initial gray value of the pixel, D is the difference parameter after pixel correction, D α is the preset allowable difference parameter, R is the preset constant, and exp is the natural exponential function.

[0066] In this embodiment, the photovoltaic array grayscale image is enhanced or suppressed by the size of the difference parameter to achieve image enhancement.

[0067] S103, obtaining grayscale values ​​of pixel points of the photovoltaic array grayscale image after image enhancement, and clustering the pixel points according to the grayscale values ​​of the pixel points of the photovoltaic array grayscale image after image enhancement;

[0068] In some embodiments of the present application, the pixel points are clustered according to the grayscale values ​​of the pixel points of the photovoltaic array grayscale image after image enhancement, including: establishing a grayscale data set according to the grayscale values ​​of the pixel points of the photovoltaic array grayscale image after image enhancement, and randomly selecting k initial clustering centers of the grayscale data set; calculating the Euclidean distance from the grayscale value in the grayscale data set to the initial clustering center, and dividing each pixel point into a corresponding cluster cluster according to the Euclidean distance from the grayscale value in the grayscale data set to the initial clustering center; calculating the grayscale average value within each cluster cluster, and re-determining the cluster center according to the grayscale average value within each cluster cluster; repeatedly iterating the above steps until the cluster center no longer changes or the number of iterations reaches a preset iteration threshold, obtaining k final cluster centers, and determining the cluster cluster to which each pixel point belongs according to the clustering result.

[0069] In this embodiment, the pixels after image enhancement are clustered based on the k-means clustering algorithm, and each pixel is clustered into a corresponding cluster according to the gray value, so as to facilitate subsequent abnormal detection of the pixel.

[0070] S104, extracting the sub-array image with abnormalities according to the clustering of the pixel points, performing edge detection on the photovoltaic panel in the sub-array image with abnormalities, and extracting the abnormal edge region of the photovoltaic panel according to the edge detection result;

[0071] In some embodiments of the present application, the extracting of abnormal sub-array images based on clustering clusters of pixel points includes: obtaining a preset grayscale abnormality threshold, calculating the difference between the grayscale value of the cluster center of each cluster and the preset grayscale abnormality threshold; if the difference between the grayscale value of the cluster center and the preset grayscale abnormality threshold is less than a first preset threshold, then judging that the sub-array image corresponding to the cluster cluster has an abnormality; if the difference between the grayscale value of the cluster center and the preset grayscale abnormality threshold is greater than or equal to the first preset threshold, then judging the difference between the average grayscale value of the sub-array image and the average grayscale value of all adjacent sub-array images; if the difference between the average grayscale value of the sub-array image and the average grayscale value of all adjacent sub-array images is less than a second preset threshold, then judging that the corresponding sub-array image has an abnormality.

[0072] In this embodiment, if the difference between the grayscale value of the cluster center of the cluster and the preset grayscale abnormality threshold is less than the first preset threshold, indicating that the pixel points in the cluster are dark, then the sub-array image to which it belongs has an abnormality; if the difference between the average grayscale value of the sub-array image and the average grayscale value of all adjacent sub-array images is less than the second preset threshold, indicating that the sub-array image and the average grayscale value of all adjacent sub-array images differ too much, then it is judged that the corresponding sub-array image has an abnormality. The clustering is used to make a preliminary judgment on the abnormality of the sub-array image, without the need to inspect the photovoltaic panels one by one, thereby effectively improving the inspection efficiency of the photovoltaic panels.

[0073] In some embodiments of the present application, the edge detection is performed on the photovoltaic panel in the sub-array image with the abnormality, and the abnormal edge area of ​​the photovoltaic panel is extracted according to the edge detection result, including: performing edge detection on the photovoltaic panel in the sub-array image with the abnormality based on the canny edge detection algorithm to obtain the edge line of the photovoltaic panel; obtaining the preset standard edge line, and calculating the matching degree between the preset standard edge line and the edge line of the photovoltaic panel; screening out the edge area where the matching degree between the preset standard edge line and the edge line of the photovoltaic panel is less than a third preset threshold, to obtain the abnormal edge area of ​​the photovoltaic panel.

[0074] In this embodiment, by performing edge detection on the photovoltaic panel in the sub-array image with abnormalities, the abnormal edge area is detected, which facilitates the subsequent determination of the defect type of the photovoltaic panel.

[0075] S105, calculating the defect characteristic value of the abnormal edge area, and determining the defect type of the photovoltaic panel according to the defect characteristic value of the abnormal edge area.

[0076] In some embodiments of the present application, the calculation of the defect characteristic value of the abnormal edge area includes: obtaining the average grayscale value of the abnormal edge area of ​​the photovoltaic panel, calculating the difference between the average grayscale value of the remaining areas of the photovoltaic panel and the average grayscale value of the abnormal edge area, and obtaining a first abnormal parameter; setting the inverse of the degree of matching between the edge line of the abnormal edge area and the standard edge line as the second abnormal parameter, calculating the product of the first abnormal parameter and the second abnormal parameter, and obtaining the defect characteristic value of the abnormal edge area.

[0077] In this embodiment, the first abnormal parameter of the abnormal edge area is determined by the difference between the average grayscale value of the remaining areas of the photovoltaic panel and the average grayscale value of the abnormal edge area, and the second abnormal parameter is determined by the degree of matching between the edge line of the abnormal edge area and the standard edge line, so that the defect characteristic value of the abnormal edge area is obtained by calculating the product of the first abnormal parameter and the second abnormal parameter.

[0078] In some embodiments of the present application, determining the defect type of the photovoltaic panel based on the defect characteristic value of the abnormal edge area includes: obtaining a preset allowable characteristic value, and calculating the difference between the defect characteristic value of the abnormal edge area and the preset allowable characteristic value; if the difference between the defect characteristic value of the abnormal edge area and the preset allowable characteristic value is greater than a fourth preset threshold, then judging that the defect type of the photovoltaic panel corresponding to the abnormal edge area is a crack defect; if the difference between the defect characteristic value of the abnormal edge area and the preset allowable characteristic value is less than or equal to the fourth preset threshold, then judging that the defect type of the photovoltaic panel corresponding to the abnormal edge area is an occlusion defect.

[0079] In this embodiment, if the difference between the defect characteristic value of the abnormal edge area and the preset allowable characteristic value is greater than the fourth preset threshold, indicating that the grayscale value of the abnormal edge area is significantly darker than that of the rest of the area, and the matching degree between the edge line of the abnormal edge area and the standard edge line is too low, then the photovoltaic panel defect type corresponding to the abnormal edge area is judged to be a crack defect; if the difference between the defect characteristic value of the abnormal edge area and the preset allowable characteristic value is less than or equal to the fourth preset threshold, then the photovoltaic panel defect type corresponding to the abnormal edge area is judged to be an occlusion defect.

[0080] Based on the same technical concept, such as Figure 2 As shown, the present invention also provides a photovoltaic panel defect inspection system based on a drone, including: a first module, used to obtain image data of a photovoltaic array, grayscale the image data of the photovoltaic array, and obtain a photovoltaic array grayscale image; a second module, used to evenly divide the photovoltaic array grayscale image to obtain a plurality of sub-array images, and perform image enhancement on the photovoltaic array grayscale image according to the sub-array images; a third module, used to obtain the grayscale values ​​of the pixel points of the photovoltaic array grayscale image after image enhancement, and cluster the pixel points according to the grayscale values ​​of the pixel points of the photovoltaic array grayscale image after image enhancement; a fourth module, used to extract the sub-array image with abnormalities according to the clustering clusters of the pixel points, perform edge detection on the photovoltaic panel in the sub-array image with abnormalities, and extract the abnormal edge area of ​​the photovoltaic panel according to the edge detection result; a fifth module, used to calculate the defect characteristic value of the abnormal edge area, and determine the defect type of the photovoltaic panel according to the defect characteristic value of the abnormal edge area.

[0081] By applying the above technical scheme, the present invention obtains the image data of the photovoltaic array, performs grayscale processing on the image data of the photovoltaic array, and obtains the photovoltaic array grayscale image; averagely divides the photovoltaic array grayscale image to obtain a number of sub-array images, and performs image enhancement on the photovoltaic array grayscale image according to the sub-array images; obtains the grayscale value of the pixel point of the photovoltaic array grayscale image after image enhancement, and clusters the pixel points according to the grayscale value of the pixel point of the photovoltaic array grayscale image after image enhancement; extracts the sub-array image with abnormalities according to the clustering cluster of the pixel points, performs edge detection on the photovoltaic panel in the sub-array image with abnormalities, and extracts the abnormal edge area of ​​the photovoltaic panel according to the edge detection result; calculates the defect characteristic value of the abnormal edge area, and determines the defect type of the photovoltaic panel according to the defect characteristic value of the abnormal edge area. It can effectively increase the accuracy of photovoltaic panel defect inspection and improve inspection efficiency.

[0082] Through the description of the above implementation methods, those skilled in the art can clearly understand that the present invention can be implemented by hardware, or by means of software plus a necessary general hardware platform. Based on such an understanding, the technical solution of the present invention can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (which can be a CD-ROM, a USB flash drive, a mobile hard disk, etc.), including a number of instructions for a computer device (which can be a personal computer, a server, or a network device, etc.) to execute the methods described in each implementation scenario of the present invention.

[0083] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present application, rather than to limit it. Although the present application has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that they can still modify the technical solutions described in the aforementioned embodiments, or make equivalent replacements for some of the technical features therein. However, these modifications or replacements do not cause the essence of the corresponding technical solutions to deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A photovoltaic panel defect inspection method based on drone, characterized in that: The method comprises: Acquire image data of the photovoltaic array, perform grayscale processing on the image data of the photovoltaic array, and obtain a grayscale image of the photovoltaic array; The photovoltaic array grayscale image is evenly divided to obtain a number of sub-array images, and the photovoltaic array grayscale image is enhanced according to the sub-array images; Obtaining grayscale values ​​of pixel points of the photovoltaic array grayscale image after image enhancement, and clustering the pixel points according to the grayscale values ​​of the pixel points of the photovoltaic array grayscale image after image enhancement; Extracting the sub-array image with abnormalities according to the clustering clusters of pixel points, performing edge detection on the photovoltaic panel in the sub-array image with abnormalities, and extracting the abnormal edge area of ​​the photovoltaic panel according to the edge detection result; The defect characteristic value of the abnormal edge area is calculated, and the defect type of the photovoltaic panel is determined according to the defect characteristic value of the abnormal edge area.

2. The photovoltaic panel defect inspection method based on drone according to claim 1 is characterized in that: The step of performing image enhancement on the photovoltaic array grayscale image according to the sub-array image comprises: Obtaining the grayscale value of the pixel in the subarray image, calculating the average grayscale value of the subarray image, and determining the difference parameter of the pixel according to the difference between the grayscale value of the pixel in the subarray image and the average grayscale value; Obtain solar radiation data corresponding to the photovoltaic array grayscale image, and determine the shadow correction coefficient of the pixel point at each position in the photovoltaic array grayscale image according to the solar radiation data; The difference parameter of the pixel point is multiplied by the shadow correction coefficient to obtain the corrected difference parameter, and the photovoltaic array grayscale image is enhanced according to the corrected difference parameter.

3. The photovoltaic panel defect inspection method based on drone according to claim 2 is characterized in that: Determining the shadow correction coefficient of each pixel point in the photovoltaic array grayscale image according to the solar radiation data includes: Determine the solar irradiation angle at each position of the photovoltaic array according to the solar irradiation data, set the initial irradiation area according to the solar irradiation angle, and determine the distance parameter value of the pixel point according to the distance between the pixel point in the photovoltaic array and the initial irradiation area; Determine the light intensity value according to the solar radiation data, and set the angle parameter value according to the light intensity value; The angle parameter value is used as the hypotenuse angle, the distance parameter value of the pixel point is converted into the hypotenuse length according to the preset ratio, and a right triangle is constructed according to the hypotenuse angle and the maximum hypotenuse length; Take the non-right-angled point of the right triangle close to the base as the initial point, determine the position of each pixel on the hypotenuse of the right triangle according to the length of the hypotenuse of each pixel, determine the height value of each pixel according to the position of each pixel on the hypotenuse of the right triangle, and determine the shadow correction coefficient according to the height value of each pixel.

4. The photovoltaic panel defect inspection method based on drone according to claim 2 is characterized in that: The step of performing image enhancement on the photovoltaic array grayscale image according to the corrected difference parameter comprises: Based on the image enhancement formula, the grayscale value of the pixel point of the photovoltaic array grayscale image is enhanced or suppressed according to the corrected difference parameter. The image enhancement formula is specifically: Among them, G is the gray value of the pixel after enhancement or suppression, G α is the initial gray value of the pixel, D is the difference parameter after pixel correction, D α is the preset allowable difference parameter, R is the preset constant, and exp is the natural exponential function.

5. The photovoltaic panel defect inspection method based on drone according to claim 1 is characterized in that: The clustering of pixels according to the grayscale values ​​of the pixels of the photovoltaic array grayscale image after image enhancement includes: A grayscale data set is established according to the grayscale values ​​of the pixels of the photovoltaic array grayscale image after image enhancement, and k initial clustering centers of the grayscale data set are randomly selected; Calculate the Euclidean distance from the grayscale value in the grayscale data set to the initial cluster center, and divide each pixel point into the corresponding cluster cluster according to the Euclidean distance from the grayscale value in the grayscale data set to the initial cluster center; Calculate the grayscale average value in each cluster, and re-determine the cluster center based on the grayscale average value in each cluster; Repeat the above steps until the cluster center no longer changes or the number of iterations reaches the preset iteration threshold, and obtain k final cluster centers. According to the clustering results, determine the cluster to which each pixel belongs.

6. The photovoltaic panel defect inspection method based on drone according to claim 5 is characterized in that: The step of extracting the sub-array image with abnormalities according to the clustering of pixel points includes: Obtain a preset grayscale anomaly threshold, and calculate the difference between the grayscale value of the cluster center of each cluster and the preset grayscale anomaly threshold; If the difference between the gray value of the cluster center and the preset gray abnormality threshold is less than the first preset threshold, it is determined that the subarray image corresponding to the cluster cluster has an abnormality; If the difference between the grayscale value of the cluster center and the preset grayscale abnormality threshold is greater than or equal to the first preset threshold, the difference between the average grayscale value of the subarray image and the average grayscale value of all adjacent subarray images is determined; if the difference between the average grayscale value of the subarray image and the average grayscale value of all adjacent subarray images is less than the second preset threshold, it is determined that the corresponding subarray image has an abnormality.

7. The photovoltaic panel defect inspection method based on drone according to claim 1 is characterized in that: The step of performing edge detection on the photovoltaic panel in the sub-array image with abnormalities and extracting the abnormal edge region of the photovoltaic panel according to the edge detection result includes: Based on the canny edge detection algorithm, the edge of the photovoltaic panel in the sub-array image with abnormality is detected to obtain the edge line of the photovoltaic panel; Obtaining a preset standard edge line, and calculating the matching degree between the preset standard edge line and the edge line of the photovoltaic panel; The edge area where the matching degree between the preset standard edge line and the edge line of the photovoltaic panel is less than the third preset threshold is screened out to obtain the abnormal edge area of ​​the photovoltaic panel.

8. The photovoltaic panel defect inspection method based on drone according to claim 7 is characterized in that: The step of calculating the defect characteristic value of the abnormal edge area includes: Obtaining an average grayscale value of an abnormal edge region of the photovoltaic panel, calculating a difference between an average grayscale value of the remaining region of the photovoltaic panel and an average grayscale value of the abnormal edge region, and obtaining a first abnormal parameter; The inverse of the matching degree between the edge line of the abnormal edge area and the standard edge line is set as the second abnormal parameter, and the product of the first abnormal parameter and the second abnormal parameter is calculated to obtain the defect characteristic value of the abnormal edge area.

9. The photovoltaic panel defect inspection method based on drone according to claim 8 is characterized in that: The method of determining the defect type of the photovoltaic panel according to the defect characteristic value of the abnormal edge area includes: Obtaining a preset allowable characteristic value, and calculating the difference between the defect characteristic value of the abnormal edge area and the preset allowable characteristic value; If the difference between the defect characteristic value of the abnormal edge area and the preset allowable characteristic value is greater than a fourth preset threshold, it is determined that the defect type of the photovoltaic panel corresponding to the abnormal edge area is a crack defect; If the difference between the defect characteristic value of the abnormal edge area and the preset allowable characteristic value is less than or equal to the fourth preset threshold, it is determined that the photovoltaic panel defect type corresponding to the abnormal edge area is an occlusion defect.

10. A photovoltaic panel defect inspection system based on drones, characterized in that: include: The first module is used to obtain the image data of the photovoltaic array, perform grayscale processing on the image data of the photovoltaic array, and obtain a grayscale image of the photovoltaic array; The second module is used to evenly divide the photovoltaic array grayscale image to obtain a number of sub-array images, and perform image enhancement on the photovoltaic array grayscale image according to the sub-array images; The third module is used to obtain the grayscale values ​​of the pixels of the photovoltaic array grayscale image after the image enhancement, and cluster the pixels according to the grayscale values ​​of the pixels of the photovoltaic array grayscale image after the image enhancement; The fourth module is used to extract the sub-array image with abnormalities according to the clustering clusters of pixel points, perform edge detection on the photovoltaic panel in the sub-array image with abnormalities, and extract the abnormal edge area of ​​the photovoltaic panel according to the edge detection result; The fifth module is used to calculate the defect characteristic value of the abnormal edge area and determine the defect type of the photovoltaic panel according to the defect characteristic value of the abnormal edge area.

Images