Insulator ceramic chip damage degree evaluation method and device based on unmanned aerial vehicle inspection

Abstract

The invention discloses an insulator ceramic chip damage degree evaluation method and device based on unmanned aerial vehicle inspection, and belongs to the technical field of image processing artificial intelligence, and the method comprises the steps: S1, collecting an insulator image in operation; s2, inputting the insulator image into a deep learning model to segment an insulator damaged ceramic chip image, and obtaining the total number of all pixels of the insulator damaged ceramic chip; s3, extracting edge points of the image of the damaged ceramic chip of the insulator to obtain the contour of the damaged ceramic chip of the insulator; s4, detecting an elliptic arc existing in the contour of the damaged ceramic chip of the insulator, and obtaining an elliptic parameter; s5, fitting the contour of the damaged ceramic chip of the insulator by using the ellipse parameter; s6, calculating the number of pixels of the incomplete part of the damaged ceramic chip of the insulator; and S7, calculating the ratio of the number of the pixels of the incomplete part of the damaged ceramic chip of the insulator to the total number of all the pixels of the insulator, and evaluating the damage degree of the ceramic chip of the insulator according to the ratio. According to the invention, the damaged area can be identified and the damage degree information can be obtained.

Description

Technical Field

[0001] The present invention belongs to the technical field of artificial intelligence for image processing, and particularly relates to a method and device for evaluating the damage degree of insulator porcelain chips based on drone inspection, which can be used for the drone inspection work of insulator performance in the power system to improve the operation and maintenance level of insulators. Background Art

[0002] With the development of the power system and the expansion of the scale of transmission lines, insulators, as an important part of the power system, their performance status is crucial for the safe operation of transmission lines. During long-term outdoor use, due to factors such as sunlight, electric shock, and lightning strike, some porcelain chips of insulators are missing, resulting in a decline in their insulation performance, which brings greater risks to the safe operation and maintenance of the power system. Therefore, it is necessary to regularly inspect insulators and repair or replace them when necessary to ensure the safe operation of the power system.

[0003] The size of the insulator damage determines its maintenance measures. Smaller damage can be repaired to extend the service life of the insulator, while larger damage cannot be repaired and must be replaced in a timely manner. Therefore, in the insulator inspection work, it is of great significance to evaluate the damage degree when damage is detected. However, current insulator damage detection only focuses on the research of methods aiming at detecting damage, and no research methods for the damage degree of insulators have been publicly reported. Therefore, the obtained damage information is limited.

[0004] Guangdong Power Grid Co., Ltd. and Electric Power Research Institute of Guangdong Power Grid Co., Ltd. disclose "A Detection Method and System for Damage of Hybrid Insulator Porcelain Parts" in the patent application document with the application publication number of CN118465069A. The method includes: knocking on the hybrid insulator to be tested and collecting the vibration signal generated by the hybrid insulator to be tested during knocking; amplifying the vibration signal and performing fast Fourier transform processing to extract the characteristic parameters of the processed vibration signal; comparing the characteristic parameters with the preset parameters, and judging whether there is a fault in the hybrid insulator to be tested according to whether the difference exceeds the difference preset threshold. This invention patent application does not adopt the method of image vision, but adopts the vibration signal analysis method. This invention can only give information on whether there is damage, cannot give the damaged area, nor can it give the damage degree information.

[0005] North China Electric Power University discloses "A method for detecting damage of composite insulators, terminal equipment and readable storage medium" in the patent application document with the publication number CN113643234A. This patent application provides a method for detecting damage of composite insulators, which includes acquiring an image of the composite insulator and preprocessing the image; performing edge detection on the preprocessed image to obtain the edge of the umbrella skirt of the composite insulator; performing ellipse detection on the edge; and detecting damage of the composite insulator based on the distance from the pixel points inside the ellipse to the ellipse edge. It can effectively identify the damage of the composite insulator. This patent uses a traditional edge detection algorithm to identify the insulator, and then determines that a point is a damaged point if the distance between the ellipse detected and the insulator edge exceeds a certain value. When the number of damaged points exceeds 100, it is determined that the insulator is damaged. It can be seen that the identification of this patent is based on traditional methods, aiming to detect damage, but the degree of damage cannot be obtained. Moreover, the traditional method using edge detection is vulnerable to environmental interference and has insufficient detection performance.

[0006] Xi'an Polytechnic University discloses "A method for detecting damage of insulators based on ellipse feature fitting" in the patent document with the authorization announcement number CN109785285B. This patent first grayscales the original image of the insulator and performs image filtering to remove the interference noise of the image, then performs two-dimensional OTSU threshold segmentation on the image to obtain the global threshold, gets the insulator area, and fills the "holes" and removes false targets from the image after two-dimensional OTSU threshold segmentation by combining morphological filtering and connected region marking; performs edge detection on the processed image to obtain the edge contour of the insulator, solves the center coordinate point and the long axis rotation angle through optimal ellipse fitting, gets the fitting ellipse of each insulator in the insulator string, and obtains the optimal ellipse fitting of the entire insulator string by analyzing the insulator pieces above and below the insulator string; finally uses the slope model for damage detection. It can be seen that this patent also uses traditional image processing methods, aiming to detect damage, but the information on the degree of damage cannot be obtained. Summary of the Invention

[0007] The purpose of the present invention is to propose a method and device for evaluating the damage degree of insulator porcelain pieces based on drone inspection for the deficiencies of the above-mentioned existing technologies, which can identify the damaged area and obtain the information on the degree of damage.

[0008] To achieve the above purpose, the present invention adopts the following technical solutions: In the first aspect, the present invention provides a method for evaluating the damage degree of insulator porcelain pieces based on drone inspection, including: S1. A power inspection drone collects images of insulators in operation on a transmission tower; S2. Input the collected insulator image into the first part of the deep learning model, DAMO-YOLO, for object detection. If the network detects a damaged area in the image, output a rectangular box to enclose the damaged area; if no damage is detected, output the detection result that the insulator is intact without damage, and the process ends. Input the rectangular box area into the second part of the deep learning model, SAM, for object pixel-level image segmentation, segment out the damaged porcelain piece image of the insulator, extract all the pixels of the damaged porcelain piece of the insulator, and obtain the total number of all pixels of the damaged porcelain piece of the insulator. S3. Extract the edge points of the damaged porcelain piece image of the insulator, and obtain the contour of the damaged porcelain piece of the insulator based on the edge points of the damaged porcelain piece image of the insulator. S4. Detect the elliptical arcs existing in the contour of the damaged porcelain piece of the insulator, and obtain the elliptical parameters. S5. Use the elliptical parameters to fit the contour of the damaged porcelain piece of the insulator. S6. Calculate the number of pixels of the incomplete part of the damaged porcelain piece of the insulator. S7. Based on the contour of the damaged porcelain piece of the insulator obtained by fitting, calculate the ratio of the number of pixels of the incomplete part of the damaged porcelain piece of the insulator to the total number of all pixels of the insulator, and evaluate the damage degree of the porcelain piece of the insulator according to the ratio.

[0009] Further, the DAMO-YOLO network includes a backbone network, a neck, and a head. The backbone network uses the CSPDarknet53 convolutional neural network model to extract image features. The neck includes a feature pyramid network and an attention mechanism module to enhance the features of key information. The head includes several modules such as a decoupled head, dynamic label assignment, and a loss function to generate the detection results of the target, including the category and bounding box of the target.

[0010] Further, the SAM network includes an image encoder, a prompt encoder, and a mask decoder. The image encoder uses the Vision Transformer backbone network to extract high-dimensional features and global semantic information of the image. The prompt encoder uses the CLIP text encoding model to encode the user prompt information into a feature vector. The mask decoder uses the Transformer decoder to generate a segmentation mask through upsampling and per-pixel prediction according to the image features and the prompt feature vector, and obtain the segmentation area of the target.

[0011] Further, the process of extracting the edge points of the damaged porcelain piece image of the insulator in step S3 is as follows: Use the Sobel operator to calculate the gradient magnitude and gradient direction of each pixel in the damaged porcelain piece image of the insulator, and the calculation formula is as follows: ; ; in, is the horizontal gradient, is the vertical gradient,

[0012] Non-maximum suppression is performed to quantize the gradient direction of each pixel to one of four angles: 0°, 45°, 90°, and 135°. The quantization rule is: quantize to the direction closest to the pixel. After quantization, the gradient of the current pixel is compared with the gradient of the two adjacent pixels along the gradient direction. If the gradient of the current pixel is the largest, it is retained; otherwise, its gradient is reset to 0. After non-maximum suppression, only the locally maximum gradients are retained. The pixels corresponding to these locally maximum gradients are edge points.

[0013] Furthermore, the Huffman ellipse detection algorithm is used in step S4 to detect the elliptical arc in the outline of the damaged insulator tile, including the following steps: S4.1. Constructing a five-dimensional parameter space ,in is the horizontal coordinate of the center point of the ellipse, is the ordinate of the center point of the ellipse, is the semi-major axis of the ellipse, is the semi-minor axis of the ellipse, is the tilt angle of the ellipse; S4.2. Randomly select 5 edge points and use the least squares method to fit the ellipse parameters , voting on the fitted ellipse parameters in parameter space; Repeat S4.2 above until the maximum number of iterations is reached and the ellipse parameter with the most votes is obtained. , the ellipse parameter with the most votes The characterized ellipse is the detected ellipse.

[0014] Furthermore, step S6 includes the following steps: Calculate the total number of pixels inside the ellipse based on the ellipse parameters, then calculate the number of pixels in the area where the tile is blocked based on the center position of the ellipse. Subtract the number of pixels in the blocked part of the tile and the total number of pixels of the damaged tile from the total number of pixels in the ellipse to get the number of pixels in the damaged part of the tile. The method for calculating the number of pixels in the area where the tile is blocked is as follows: Set it as a seed point, and then use the region growing algorithm to obtain the region where the occluded part is located, count the pixels in the region where the occluded part is located, and obtain the number of pixels in the region where the occluded part of the porcelain chip is located.

[0015] Further, in step 7, the method for evaluating the damage degree of the insulator porcelain chip includes: Calculate the damage ratio L through the following formula, , where M is the number of pixels in the incomplete part of the damaged porcelain chip, and X is the total number of all pixels of the damaged insulator porcelain chip; Damage degree = .

[0016] In a second aspect, the present invention provides an apparatus for estimating the damage degree of insulator porcelain chips based on drone inspection, including: An acquisition module for acquiring images of insulators in operation on a transmission tower; A porcelain chip damage detection module for detecting damage to porcelain chips based on images of insulators in operation; A damaged porcelain chip contour extraction module for extracting edge points of an image of a damaged insulator porcelain chip and obtaining the contour of the damaged insulator porcelain chip; An ellipse detection module for detecting elliptical arcs existing in the contour of the damaged insulator porcelain chip and obtaining ellipse parameters; An ellipse contour fitting module for fitting the contour of the damaged insulator porcelain chip by using the ellipse parameters; A porcelain chip incomplete pixel number calculation module for calculating the number of pixels in the incomplete part of the damaged insulator porcelain chip; A damage degree evaluation module for calculating the ratio of the number of pixels in the incomplete part of the damaged insulator porcelain chip to the total number of all pixels of the insulator based on the fitted contour of the damaged insulator porcelain chip, and evaluating the damage degree of the insulator porcelain chip according to the ratio.

[0017] In a third aspect, the present invention provides an electronic device, including: At least one processor; and, A memory communicatively connected to the at least one processor; wherein, The memory stores instructions executable by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the method for evaluating the damage degree of insulator porcelain chips based on drone inspection according to any one of the first aspects of the present invention.

[0018] In a fourth aspect, the present invention provides a computer-readable storage medium storing a computer program, and when the computer program is executed by a processor, it implements the method for evaluating the damage degree of insulator porcelain chips based on drone inspection according to any one of the first aspects of the present invention.

[0019] Compared with the prior art, the present invention has at least the following beneficial technical effects: This paper uses the DAMO-YOLO target detection network and the SAM semantic segmentation network fusion mechanism to construct a deep learning model that simultaneously detects insulator damage and segments damaged tiles. This model can not only effectively detect damaged insulator areas, but also segment the specific semantic information of the damaged areas. This makes it easier to obtain information on the degree of damage and provide more detailed information on damaged tiles. The degree of damage is assessed by comparing the number of pixels in the missing part of the damaged tile to the total number of pixels on the insulator, providing more sufficient data support for specific maintenance decisions. This model is suitable for use in situations where drones are used to inspect the status of insulators.

[0020] The present invention uses ellipse fitting technology to compare with the damaged tile segmentation result obtained by the present invention to obtain the damaged area of the tile, and on this basis, provides a method for estimating the degree of damage, which is convenient for taking different operation and maintenance measures. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Figure 1 It is a model block diagram of the present invention; Figure 2 It is the insulator processing flow chart; Figure 3 It is a schematic diagram of the damaged area; Figure 4 A structural block diagram of an insulator porcelain damage degree estimation device based on drone inspection provided by an embodiment of the present invention; Figure 5 A block diagram of an electronic device according to an embodiment of the present invention. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the solutions of the present invention, the technical solutions in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the embodiments described are only 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 ordinary technicians in this field without making creative efforts should fall within the scope of protection of the present invention.

[0023] It should be noted that the terms "first", "second", etc. in the description, claims and above-mentioned drawings of the present invention are used to distinguish similar objects, and do not necessarily have to be used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged under appropriate circumstances, so that the embodiments of the present invention described here can be implemented in an order other than those illustrated or described here. In addition, the terms "comprising" and "having" and any variations thereof are intended to cover non-exclusive inclusion. For example, a process, method, system, product or device comprising a series of steps or units does not necessarily have to be limited to those steps or units clearly listed, but may include other steps or units not clearly listed or inherent to these processes, methods, products or devices.

[0024] The method for evaluating the damage degree of insulator porcelain chips based on UAV inspection provided by the present invention serves the scenario of UAV inspection of insulator states. The method includes collecting insulator images by UAV, and then applying a two-stage deep learning model of DAMO-YOLO+SAM for object detection and semantic segmentation to the collected images to detect damaged porcelain chips and the semantic segmentation regions of the damaged porcelain chips. Then, through ellipse detection and fitting, the fitting complete contour of the damaged porcelain chips is obtained. The fitting contour is compared with the contour of the segmentation region, and the actual occlusion region is removed to obtain the damaged region of the porcelain chips. Finally, according to the ratio of the number of pixels in the damaged region to the number of pixels of the semantic segmentation of the damaged porcelain chips itself, the damage degree of the porcelain chips is determined, which is divided into four grades: excellent, good, medium and poor.

[0025] The present invention will be further described in detail below with reference to the accompanying drawings: Embodiment 1 Refer to Figures 1 to 3 , this embodiment provides a method for evaluating the damage degree of insulator porcelain chips based on UAV inspection, including the following steps: Step 1. Use a power inspection UAV to collect images of insulators in operation on a transmission tower; Step 2. Input the collected insulator images into a deep learning model to detect the damaged targets of insulator porcelain chips; Step 2.1 The collected insulator images are first input into the first part of the deep learning model, DAMO-YOLO, for object detection. If the DAMO-YOLO network detects a damaged region in the image, it outputs a rectangular box to enclose the damaged region. The coordinate parameters of the rectangular box are (x, y, w, h), where x is the abscissa of the lower left corner point of the rectangular box, y is the ordinate of the lower left corner point of the rectangular box, w is the width of the rectangular box, and h is the height of the rectangular box; The DAMO-YOLO network is a deep learning object detection network, which consists of three parts: the backbone network, the neck, and the head. The backbone network uses the CSPDarknet53 convolutional neural network model to extract image features. The neck includes the Feature Pyramid Network (FPN) and the attention mechanism module to enhance the features of key information. The head includes modules such as the Decoupled Head, dynamic label assignment, and loss function to generate the detection results of the target, including the category and bounding box of the target. Step 2.2 Then input the rectangular box area (x, y, w, h) output in Step 2.1 into the second part of the deep learning model, the SAM model, for target pixel-level image segmentation to segment the damaged porcelain chip image, identify and extract all the pixels of the damaged porcelain chip of the insulator, and record the total number of all pixels as X , and the pixel values of the remaining non-damaged porcelain chips of the insulator are set to 0; the damaged porcelain chip of the insulator is the porcelain chip of the insulator with a damaged area. The SAM network is a deep learning image segmentation network, including the following three modules: the image encoder, the prompt encoder, and the mask decoder. The image encoder uses the Vision Transformer backbone network to extract high-dimensional features and global semantic information of the image. The prompt encoder uses the CLIP text encoding model to encode the user prompt information into a feature vector. The mask decoder uses the Transformer decoder to generate a segmentation mask through upsampling and per-pixel prediction according to the image features and the prompt feature vector to obtain the segmentation area of the target. Step 2.3 If no damage is detected in Step 2.1, output the information that the insulator is intact without damage, and the process ends. Step 3. Perform an edge extraction operation on the damaged porcelain chip image of the insulator extracted in Step 2.2 to obtain the contour information of the damaged porcelain chip of the insulator. For the edge extraction operation, the canny edge detection algorithm is used. The process is as follows: Use the Sobel operator to calculate the gradient of each pixel in the damaged porcelain chip image of the insulator , including calculating the gradient magnitude and gradient direction. The calculation formulas are as follows: Horizontal gradient , vertical gradient , (1) Gradient magnitude , (2) Gradient direction , (4) Then, non-maximum suppression is performed to quantize the gradient direction of each pixel to one of the four angles (0º, 45º, 90º, 135º). The quantization rule is to quantize to the direction closest to the pixel. After quantization, the gradient values of the current pixel and the two adjacent pixels are compared along the gradient direction. If the gradient of the current pixel is the largest, it is retained; otherwise, its gradient is reset to 0. After non-maximum suppression, only the local maximum gradients are retained. The pixels corresponding to these local maximum gradients are edge points, and all edge points form the contour.

[0026] Step 4. Use the Hough ellipse detection algorithm to detect the elliptical arcs in the contour of the edge points extracted in step 3 and obtain the ellipse parameters; The process of Huffman ellipse detection algorithm is as follows: Step 4.1 First construct a five-dimensional parameter vector ,in is the horizontal coordinate of the center point of the ellipse, is the ordinate of the center point of the ellipse, is the semi-major axis of the ellipse, is the semi-minor axis of the ellipse, is the tilt angle of the ellipse; Step 4.2 Next, randomly select 5 points from the edge points and use the least squares method to fit the ellipse parameters. , voting on the fitted ellipse parameters in parameter space; Repeat step 4.2 until the maximum number of iterations is reached and the ellipse parameter with the most votes is obtained. , which is the detected ellipse. At this time, the edge points that have voted for this parameter are all points on the ellipse; Step 5. Use the ellipse parameters obtained in step 4 to fit the ellipse contour of the damaged insulator porcelain piece; that is, use the ellipse parameters Draw the entire oval.

[0027] Step 6. Calculate the number of pixels of the damaged part of the insulator, that is, the number of pixels between the outline of the damaged part and the fitted ellipse outline; Calculate the number of pixels of the damaged part of the insulator. The process is as follows: first calculate the total number of pixels inside the ellipse according to the ellipse parameters S , and then calculate the area where the tile is blocked according to the center position of the ellipse (i.e. Figure 3 The number of pixels in the area where the number 2 is located N Finally, the total number of pixels of the ellipse S Subtract the number of pixels of the occluded part of the tile N, and then subtract the total pixel value of the damaged porcelain pieces of the insulator obtained in the image segmentation in step 2.2 X (that is Figure 3 the total pixel value of the area where the number 1 is located in M ), to obtain the number of pixels of the incomplete part of the damaged porcelain piece of the insulator M = S - N - X , that is N , wherein the method for calculating the number of pixels of the area where the porcelain piece is blocked is: the center of the ellipse

[0028] can be set as the seed point, and then the region growth algorithm can be used to obtain the area where the blocked part is located, and the pixels of the area where the blocked part is located are counted to obtain the number of pixels N of the area where the porcelain piece is blocked. Figure 3 As shown in ; The damage degree of the insulator porcelain piece is given by the following method: Calculate the damage ratio L through the following formula , (4) Damage degree =

[0029] The following is the device embodiment of the present invention, which can be used to execute the method embodiment of the present invention. For the details not disclosed in the device embodiment, please refer to the method embodiment of the present invention.

[0030] Embodiment 2 Please refer to Figure 4 , in this embodiment, a device for estimating the damage degree of insulator porcelain pieces based on drone inspection is provided, including: An acquisition module, configured to acquire images of insulators in operation on a transmission tower; A porcelain piece damage detection module, configured to perform porcelain piece damage detection based on the images of insulators in operation; A damaged porcelain piece contour extraction module, configured to extract edge points of an image of a damaged porcelain piece of an insulator, and obtain a contour of the damaged porcelain piece of the insulator based on the edge points of the image of the damaged porcelain piece of the insulator; An ellipse detection module, configured to detect elliptical arcs existing in the contour of the damaged porcelain piece of the insulator, and obtain ellipse parameters; An ellipse contour fitting module, configured to fit a contour of the damaged porcelain piece of the insulator by using the ellipse parameters; The porcelain chip defective pixel number calculation module is used to calculate the number of defective pixels in the defective part of the damaged porcelain chip of the insulator; The damage degree evaluation module is used to calculate the ratio of the number of defective pixels in the defective part of the damaged porcelain chip of the insulator to the total number of all pixels of the insulator based on the fitted contour of the damaged porcelain chip of the insulator, and evaluate the damage degree of the porcelain chip of the insulator according to the ratio.

[0031] All relevant contents of each step involved in the embodiment of the foregoing method for evaluating the damage degree of porcelain chips of insulators based on UAV inspection can be cited in the function description of the corresponding functional modules of the device for estimating the damage degree of porcelain chips of insulators based on UAV inspection in the embodiment of the present invention, and will not be elaborated herein.

[0032] Embodiment 3 Referring to Figure 5 , this embodiment provides an electronic device, which includes a processor and a memory, and the processor is connected to the memory through a bus; the memory is used to store a computer program, the computer program includes program instructions, and the processor is used to execute the program instructions stored in the computer storage medium. The processor may be a central processing unit (CPU), or may also be other general-purpose processors, digital signal processors (DSPs), application specific integrated circuits (ASICs), field-programmable gate arrays (FPGAs) or other programmable logic devices, discrete gate or transistor logic devices, discrete hardware components, etc. It is the computing core and control core of the terminal, and is suitable for implementing one or more instructions, specifically suitable for loading and executing one or more instructions in the computer storage medium to implement the corresponding method flow or corresponding function; the processor described in the embodiment of the present invention can be used for the operation of the method for evaluating the damage degree of porcelain chips of insulators based on UAV inspection. The bus may be a peripheral component interconnect (PCI) bus or an extended industry standard architecture (EISA) bus, etc. The bus can be divided into an address bus, a data bus, a control bus, etc. For the sake of simplicity of representation, Figure 5 only one line is shown in

[0033] Embodiment 4 This embodiment provides a storage medium, specifically a computer-readable storage medium (Memory). The computer-readable storage medium is a memory device in an electronic device, used to store programs and data. It can be understood that the computer-readable storage medium here can include both the built-in storage medium in the electronic device and, of course, the extended storage medium supported by the electronic device. The computer-readable storage medium provides a storage space, and this storage space stores the operating system of the terminal. Moreover, one or more instructions suitable for being loaded and executed by the processor are stored in this storage space, and these instructions can be one or more computer programs (including program codes). It should be noted that the computer-readable storage medium here can be a high-speed RAM memory or a non-volatile memory, such as at least one disk memory. One or more instructions stored in the computer-readable storage medium can be loaded and executed by the processor to implement the corresponding steps of the method for evaluating the damage degree of insulator porcelain chips based on drone inspection in the above embodiment.

[0034] Those skilled in the art should understand that the embodiments of the present application can be provided as a method, a device, or a computer program product. Therefore, the present application can take the form of a complete hardware embodiment, a complete software embodiment, or an embodiment combining software and hardware aspects. Moreover, the present application can take the form of a computer program product implemented on one or more computer-usable storage media (including but not limited to disk memories, CD-ROMs, optical memories, etc.) containing computer-usable program codes.

[0035] The present application is described with reference to the flowcharts and / or block diagrams of methods, devices (apparatus), and computer program products according to the embodiments of the present application. It should be understood that each flow and / or block in the flowchart and / or block diagram, as well as the combination of flows and / or blocks in the flowchart and / or block diagram, can be implemented by computer program instructions. These computer program instructions can be provided to the processor of a general-purpose computer, a special-purpose computer, an embedded processor, or other programmable data processing devices to generate a machine, so that the instructions executed by the processor of the computer or other programmable data processing devices generate a device for implementing the functions specified in Figure 1 one flow or multiple flows and / or blocks Figure 1 one block or multiple blocks.

[0036] These computer program instructions can also be stored in a computer-readable memory that can direct a computer or other programmable data processing device to work in a specific manner, so that the instructions stored in the computer-readable memory generate a manufactured article including an instruction device, and this instruction device implements the functions in Figure 1 one flow or multiple flows and / or blocks Figure 1the functions specified in one or more boxes.

[0037] These computer program instructions can also be loaded onto a computer or other programmable data processing device, so that a series of operation steps are executed on the computer or other programmable device to generate a computer-implemented process. Thus, the instructions executed on the computer or other programmable device provide for implementing the steps for the functions specified in one Figure 1 one process or more processes and / or boxes Figure 1 step for the functions specified in one or more boxes.

[0038] Embodiment 5 This embodiment provides a computer program product, including a non-volatile computer-readable storage medium. The non-volatile computer-readable storage medium stores a computer program product. When the computer program is executed by a processor, it implements the steps of the methods in various embodiments of the present application.

[0039] It should be noted that the user information (including but not limited to user device information, user personal information, etc.) and data (including but not limited to data for analysis, stored data, displayed data, etc.) involved in the present application are all information and data authorized by the user or fully authorized by all parties. Moreover, the collection, use, and processing of relevant data need to comply with the relevant laws, regulations, and standards of relevant countries and regions.

[0040] In the description of this specification, the description with reference to terms such as "one embodiment", "some embodiments", "example", "specific example", or "some examples" means that the specific features, structures, materials, or characteristics described in connection with the embodiment or example are included in at least one embodiment or example of the present application. In this specification, the schematic representations of the above terms do not necessarily refer to the same embodiment or example. Moreover, the specific features, structures, materials, or characteristics described can be combined in a suitable manner in any one or more embodiments or examples. In addition, without contradiction, those skilled in the art can combine and combine the different embodiments or examples described in this specification and the features of different embodiments or examples.

[0041] Finally, it should be noted that the above embodiments are only used to illustrate the technical solutions of the present invention and not to limit them. Although the present invention has been described in detail with reference to the above embodiments, those of ordinary skill in the art should understand that: the specific implementation manners of the present invention can still be modified or equivalently replaced, and any modification or equivalent replacement that does not depart from the spirit and scope of the present invention should be covered by the protection scope of the claims of the present invention.

Claims

1. An insulator porcelain chip damage degree evaluation method based on drone inspection, characterized in that, Including: S1. The power inspection UAV collects the images of the insulators on the operating transmission towers. S2. Input the collected insulator images into the first part of the deep learning model, DAMO-YOLO, for object detection. If the network detects a damaged area of the insulator in the image, it outputs a rectangular box to enclose the damaged area. If no damage is detected, it outputs the detection result that the insulator is intact without damage, and the process ends. Input the rectangular box area into the second part of the deep learning model, SAM, for object pixel-level image segmentation, segment out the image of the damaged porcelain piece of the insulator, extract all the pixels of the damaged porcelain piece of the insulator, and obtain the total number of all pixels of the damaged porcelain piece of the insulator. S3. Extract the edge points of the image of the damaged porcelain piece of the insulator, and obtain the contour of the damaged porcelain piece of the insulator based on the edge points of the image of the damaged porcelain piece of the insulator. S4. Detect the elliptical arcs existing in the contour of the damaged porcelain piece of the insulator, and obtain the elliptical parameters. S5. Use the elliptical parameters to fit the contour of the damaged porcelain piece of the insulator. S6. Calculate the number of pixels of the incomplete part of the damaged porcelain piece of the insulator. S7. Based on the fitted contour of the damaged porcelain piece of the insulator, calculate the ratio of the number of pixels of the incomplete part of the damaged porcelain piece of the insulator to the total number of all pixels of the insulator, and evaluate the damage degree of the porcelain piece of the insulator according to the ratio.

2. The method for evaluating the damage degree of insulator porcelain chips based on UAV inspection according to claim 1, characterized in that, The DAMO-YOLO network includes a backbone network, a neck, and a head. The backbone network adopts the CSPDarknet53 convolutional neural network model to extract image features. The neck includes a feature pyramid network and an attention mechanism module to enhance the features of key information. The head includes several modules such as a decoupled head, dynamic label assignment, and a loss function to generate the detection results of the target, including the category and bounding box of the target.

3. The method for evaluating the damage degree of insulator porcelain chips based on drone inspection according to claim 1, wherein The SAM network includes an image encoder, a prompt encoder, and a mask decoder. The image encoder adopts the VisionTransformer backbone network to extract high-dimensional features and global semantic information of the image. The prompt encoder adopts the CLIP text encoding model to encode the user prompt information into a feature vector. The mask decoder adopts the Transformer decoder to generate a segmentation mask by upsampling and per-pixel prediction according to the image features and the prompt feature vector, and obtain the segmentation area of the target.

4. The method for evaluating the damage degree of insulator porcelain chips based on drone inspection according to claim 1, characterized in that The process of extracting the edge points of the damaged porcelain piece image of the insulator in step S3 is as follows: Use the Sobel operator to calculate the damaged porcelain piece image of the insulator the gradient magnitude of each pixel in and the gradient direction , and the calculation formula is as follows: ； ； Among them, is the horizontal gradient, is the vertical gradient, Perform non-maximum suppression, quantize the gradient direction of each pixel into one of the four angles of 0º, 45º, 90º, and 135º. The quantization rule is: quantize it to the direction it is closest to. After quantization, compare the gradient of the current pixel with the gradients of the two adjacent pixels along the gradient direction. If the gradient of the current pixel is the largest, keep it; otherwise, reset its gradient to 0. After non-maximum suppression, only the locally maximum gradients are retained, and the pixel points corresponding to these locally maximum gradients are the edge points.

5. The method for evaluating the damage degree of insulator porcelain chips based on UAV inspection according to claim 1, wherein, In step S4, the Huffman ellipse detection algorithm is used to detect the elliptical arcs existing in the contour of the damaged porcelain piece of the insulator, including the following steps: S4.

1. Construct a five-dimensional parameter space , where is the abscissa of the center point of the ellipse, is the ordinate of the center point of the ellipse, is the semi-major axis of the ellipse, is the semi-minor axis of the ellipse, is the tilt angle of the ellipse; S4.

2. Randomly select 5 edge points and use the least squares method to fit the ellipse parameters , and vote on the fitted ellipse parameters in the parameter space; Repeat the above S4.2 until the maximum number of iterations is reached, and obtain the ellipse parameters with the most votes. , the ellipse parameters with the most votes The ellipse characterized by the ellipse parameters is the detected ellipse.

6. The method for evaluating the damage degree of insulator porcelain chips based on drone inspection according to claim 1, wherein Step S6 includes the following steps: Calculate the total number of pixels inside the ellipse based on the ellipse parameters, then calculate the number of pixels in the area where the tile is blocked based on the center position of the ellipse. Subtract the number of pixels in the blocked part of the tile and the total number of pixels of the damaged tile from the total number of pixels in the ellipse to get the number of pixels in the damaged part of the tile. Among them, the method for calculating the number of pixels in the area where the porcelain chip is occluded is as follows: Set the center of the ellipse as the seed point, then use the region growing algorithm to obtain the area where the occluded part is located, count the pixels in the area where the occluded part is located, and obtain the number of pixels in the area where the porcelain chip is occluded.

7. The method for evaluating the damage degree of insulator porcelain chips based on drone inspection according to claim 1, wherein In step 7, evaluating the damage degree of the insulator porcelain piece includes: The damage ratio L is calculated by the following formula: , where M is the number of pixels of the damaged part of the broken porcelain chip, and X is the total number of all pixels of the damaged porcelain chip of the insulator; Degree of damage= .

8. An insulator porcelain chip damage degree estimation device based on drone inspection, characterized in that, include: An acquisition module is used to capture images of insulators in operation on transmission towers; Porcelain damage detection module, used to detect porcelain damage based on images of insulators in operation; The damaged tile contour extraction module is used to extract the edge points of the damaged insulator tile image and obtain the contour of the damaged insulator tile; The ellipse detection module is used to detect the ellipse arc in the outline of the damaged insulator porcelain piece and obtain the ellipse parameters; An ellipse contour fitting module, used to fit the contour of the damaged insulator porcelain piece using the ellipse parameters; The module for calculating the number of pixels of damaged tiles is used to calculate the number of pixels of the damaged part of the insulator tiles; The damage degree assessment module is used to calculate the ratio of the number of pixels of the damaged part of the insulator broken porcelain piece to the total number of all pixels of the insulator based on the fitted outline of the damaged insulator porcelain piece, and assess the damage degree of the insulator porcelain piece according to the ratio.

9. An electronic device, characterized in that, include: at least one processor; as well as, a memory communicatively connected to the at least one processor; wherein, The memory stores instructions that can be executed by the at least one processor, and the instructions are executed by the at least one processor so that the at least one processor can execute the insulator porcelain damage degree assessment method based on drone inspection as described in any one of claims 1 to 7.

10. A computer-readable storage medium storing a computer program, characterized in that, When the computer program is executed by a processor, the method for evaluating the degree of damage of insulator porcelain pieces based on drone inspection according to any one of claims 1 to 7 is implemented.

Images