Detection method and device for fault component of power transmission line, equipment and storage medium

Abstract

The embodiment of the invention discloses a method and device for detecting a fault component of a power transmission line, equipment and a storage medium. The method comprises the following steps: acquiring a multispectral image sequence, geographic information and meteorological information of a target power transmission line; performing image analysis on the multispectral image sequence to obtain a dynamic region image in which the target power transmission line exists; performing noise reduction analysis on the dynamic region image according to geographic information and meteorological information to obtain a component abnormal region image in the target power transmission line; and acquiring abnormal features in the abnormal region image of the component, and inputting the abnormal features into a pattern recognition algorithm to obtain a recognition result. Information of different wave bands is captured by collecting a multispectral image sequence of a power transmission line, so that image data can be effectively obtained under the condition of poor illumination conditions; and secondly, the dynamic area image is denoised in combination with meteorological information and geographic information, so that the interference of geographic and meteorological factors on the detection result can be reduced, and the detection result is more accurate.

Description

Technical Field

[0001] The present invention relates to the field of fault detection technology, and in particular to a method and device, equipment and storage medium for detecting faulty components of a power transmission line. Background Art

[0002] In the power transmission network, it is crucial to ensure the stable operation of transmission lines. As the scale of the power grid expands, the geographical coverage of transmission lines is getting wider and wider, and the environmental conditions are becoming more diverse, which increases the difficulty of line inspection and maintenance. The traditional manual line inspection method can no longer meet the requirements of modern power grids for efficiency and safety.

[0003] Existing solutions usually use multispectral cameras carried by drones to take real-time photos and analyze the captured images to check whether there are faulty components in the transmission lines. However, these methods have problems with low recognition rate and inaccurate positioning under complex environmental conditions, such as bad weather or insufficient light. Summary of the invention

[0004] Based on this, it is necessary to propose a detection method and device, equipment and storage medium for faulty components of transmission lines to address the above problems, so as to achieve accurate detection and positioning of faulty components in transmission lines.

[0005] To achieve the above-mentioned object, the present application provides, in a first aspect, a method for detecting a faulty component of a power transmission line, the method comprising:

[0006] Collecting a multispectral image sequence of a target transmission line, as well as geographic information and meteorological information of the target transmission line;

[0007] Performing image analysis on the multispectral image sequence based on differential image technology to obtain a dynamic area image where the target transmission line exists, wherein the dynamic area image is an area image where an object in the multispectral image changes;

[0008] Performing noise reduction analysis on the dynamic area image according to the geographic information and meteorological information to obtain an abnormal area image of a component in the target transmission line;

[0009] Abnormal features in the image of the abnormal area of ​​the component are collected, and the abnormal features are input into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm, wherein the recognition result is that the component in the abnormal area of ​​the component is normal or the component in the abnormal area of ​​the component is faulty.

[0010] Furthermore, the method further comprises:

[0011] Retrieving the positioning information of the component abnormal area image from the geographic information of the target transmission line, and determining the component abnormal area of ​​the target transmission line according to the positioning information;

[0012] Deploy a structured lighting device at the abnormal region of the component of the target power transmission line, and control the structured lighting device to project structured light to the abnormal region of the component according to a preset lighting mode;

[0013] The collecting of abnormal features in the image of the abnormal region of the component and inputting the abnormal features into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm specifically includes:

[0014] Acquiring a structured light image of an abnormal region of the component;

[0015] Abnormal features in the structured light image are extracted, and the abnormal features are input into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm.

[0016] Furthermore, the extracting of abnormal features from the structured light image and inputting the abnormal features into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm specifically includes:

[0017] Decoding the structured light image based on a preset image processing technology to obtain depth information of the structured light image;

[0018] enhancing the texture contrast in the structured light image according to the depth information of the structured light image to obtain an enhanced texture image;

[0019] The texture features in the texture image are extracted using a preset feature extraction algorithm, and the texture features are input into the pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm.

[0020] Furthermore, the image analysis of the multispectral image sequence based on the differential image technology to obtain the dynamic area image where the target transmission line exists specifically includes:

[0021] Selecting any two adjacent multispectral images in the multispectral image sequence, and using the multispectral images of the two adjacent frames as target multispectral images;

[0022] Preprocessing the target multispectral image to obtain a denoised target multispectral image, wherein the preprocessing includes at least normalization and / or brightness correction;

[0023] Calculating the pixel difference between the denoised target multispectral images to obtain a differential image;

[0024] The pixel area in the differential image is denoised to obtain a dynamic area image where the target power transmission line exists.

[0025] Furthermore, the denoising of the pixel area in the differential image to obtain the dynamic area image where the target power transmission line exists specifically includes:

[0026] De-noising the differential image based on a threshold processing technique to obtain a de-noised initial dynamic region image;

[0027] A morphological operation is performed on the initial dynamic region image to obtain a target dynamic region image.

[0028] Further, the performing of noise reduction analysis on the dynamic area image according to the geographic information and the meteorological information to obtain an abnormal area image of a component in the target transmission line specifically includes:

[0029] Searching for target meteorological data of the area where the dynamic area image is located from the geographic information and the meteorological information;

[0030] Analyze the dynamic impact of the target meteorological data on the non-power facilities in the dynamic area image to obtain a first false positive dynamic area in the dynamic area image, where the false positive dynamic area is an area in the dynamic area image where the non-power facilities change;

[0031] Excluding the first false positive dynamic area in the dynamic area image to obtain a standard dynamic area image;

[0032] Determining an actual line area of ​​the target transmission line in the standard dynamic area image according to the geographic information of the target transmission line;

[0033] The other regions in the standard dynamic region image except the actual line region are excluded to obtain the component abnormal region image of the target power transmission line.

[0034] Furthermore, the method further comprises:

[0035] When the identification result is a component failure within the component abnormality area, retrieving the location data of the faulty component from the geographic information of the target transmission line;

[0036] A maintenance instruction for the faulty component is generated according to the location data of the faulty component, and the maintenance instruction is outputted so as to perform maintenance on the faulty component based on the maintenance instruction.

[0037] To achieve the above-mentioned object, the second aspect of the present application provides a detection device for a faulty component of a power transmission line, the device comprising a data acquisition unit, an image analysis unit and an abnormality judgment unit;

[0038] The data acquisition unit is used to acquire a multispectral image sequence of a target power transmission line, as well as geographic information and meteorological information of the target power transmission line;

[0039] The image analysis unit is used to perform image analysis on the multispectral image sequence based on differential image technology to obtain a dynamic area image where the target transmission line exists, wherein the dynamic area image is an area image where an object in the multispectral image changes;

[0040] Performing noise reduction analysis on the dynamic area image according to the geographic information and meteorological information to obtain an abnormal area image of a component in the target transmission line;

[0041] The abnormality judgment unit is used to collect abnormal features in the image of the abnormal area of ​​the component, and input the abnormal features into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm, wherein the recognition result is that the component in the abnormal area of ​​the component is normal or the component in the abnormal area of ​​the component is faulty.

[0042] To achieve the above-mentioned purpose, the third aspect of the present application provides a computer-readable storage medium storing a computer program, characterized in that when the computer program is executed by a processor, the processor executes the steps of the method described in the first aspect.

[0043] To achieve the above-mentioned purpose, the fourth aspect of the present application provides a computer device, including a memory and a processor, characterized in that the memory stores a computer program, and when the computer program is executed by the processor, the processor executes the steps of the method described in the first aspect.

[0044] The embodiments of the present invention have the following beneficial effects:

[0045] The embodiment of the present invention proposes a method for detecting faulty components of a power transmission line, the method comprising: collecting a multispectral image sequence of a target power transmission line, as well as geographic information and meteorological information of the target power transmission line; performing image analysis on the multispectral image sequence based on differential image technology to obtain a dynamic area image of the target power transmission line, the dynamic area image being an area image where an object in the multispectral image changes; performing noise reduction analysis on the dynamic area image according to geographic information and meteorological information to obtain an abnormal area image of a component in the target power transmission line; collecting abnormal features in the abnormal area image of the component, and inputting the abnormal features into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm, the recognition result being that the component in the abnormal area of ​​the component is normal or the component in the abnormal area of ​​the component is faulty. By collecting a multispectral image sequence of a power transmission line to capture information of different bands, image data can be effectively obtained even under poor lighting conditions, thereby improving the accuracy of the detection results; secondly, by combining meteorological information and geographic information to denoise the dynamic area image, the interference of geographical and meteorological factors on the detection results can be reduced, making the detection results more accurate. BRIEF DESCRIPTION OF THE DRAWINGS

[0046] In order to more clearly illustrate the embodiments of the present invention or the technical solutions in the prior art, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative work.

[0047] in:

[0048] Figure 1 Schematic diagram of a flow chart of a method for detecting a faulty component of a power transmission line in an embodiment of the present invention;

[0049] Figure 2 A structural block diagram of a device for detecting a faulty component of a power transmission line according to an embodiment of the present invention;

[0050] Figure 3 1 is a diagram showing the internal structure of a computer device in an embodiment of the present invention. DETAILED DESCRIPTION

[0051] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are 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 creative work are within the scope of protection of the present invention.

[0052] In bad weather or insufficient light, the captured images will contain a lot of noise. If analysis is performed based on images with large noise, it will lead to large errors in the analysis results.

[0053] In the present invention, by shooting the transmission line image, fault identification and location are performed based on the transmission line image, and the typical components in the transmission line are checked for faults and the location of the faulty components is determined. If the image is shot due to factors such as bad weather or insufficient light, the recognition rate of the analysis result obtained based on the graphic analysis will be low and the positioning will not be accurate enough.

[0054] Based on this, the embodiment of the present invention proposes a method for detecting faulty components of a power transmission line to reduce the impact of climatic factors such as weather and light on the detection results, and effectively improve the accuracy of the detection results. Figure 1 , Figure 1 The figure is a flow chart of a method for detecting a faulty component of a power transmission line in an embodiment of the present invention, and the method comprises:

[0055] Step 110: Collect a multispectral image sequence of the target transmission line, as well as geographic information and meteorological information of the target transmission line.

[0056] In the embodiment of the present invention, the target transmission line is any section of the transmission line. A transmission line in a section where faults frequently occurred in the past may be selected and tested, which can greatly reduce the adverse effects of transmission line faults.

[0057] Specifically, a drone equipped with a multispectral camera is deployed into the monitoring area of ​​the target transmission line. After the drone is started, it enters flight mode and is ready to perform the monitoring task. During the flight of the drone, the multispectral camera captures images of the target transmission line and the surrounding environment at certain time intervals and trigger conditions to form a multispectral image sequence. Since multispectral images can capture information in a wider spectral range, such as bands other than visible light, such as infrared light and ultraviolet light, multispectral images contain richer information than traditional RGB images, which can enhance the image's ability to identify different materials and object states. Therefore, collecting multispectral images of the target transmission line can reduce the adverse effects of climate factors on image capture, which is conducive to detecting the state of the target transmission line under different weather conditions.

[0058] In order to ensure the quality and coverage of the image, the flight altitude and speed of the drone can be adjusted according to the actual situation. This can be done automatically by the drone's own autonomous navigation system or by manual remote control from the ground station. The purpose of the adjustment is to ensure that the collected image data is clear and comprehensive to meet the needs of the monitoring task.

[0059] In addition, the precise geographic information of the target transmission line and the meteorological information of the target transmission line location, such as wind speed, humidity and temperature, are obtained based on the geographic information system (GIS) to obtain the meteorological data when shooting multispectral images. Specifically, the precise geographic information and attitude information of all pixels in each multispectral image can be synchronously recorded based on the GPS and IMU (inertial measurement unit) data of the drone to facilitate geographic alignment between images and subsequent geographic information system integration.

[0060] By implementing the above steps, efficient and accurate monitoring of the transmission line and its surrounding environment is achieved, ensuring all-weather, all-round image acquisition capabilities. The use of multispectral imaging technology enhances the ability to identify the status of different components of the transmission line. Combined with the flexibility of drones and automated navigation technology, the image quality and monitoring coverage are improved, thereby improving the accuracy and efficiency of fault detection. At the same time, by synchronously recording the geographic location and attitude information of the image, the geographic alignment between images and the integration of geographic information systems are facilitated, providing strong technical support for subsequent fault location and maintenance, and greatly reducing the workload and safety risks of manual inspections.

[0061] Step 130: Perform image analysis on the multispectral image sequence based on differential image technology to obtain a dynamic area image where the target transmission line exists. The dynamic area image is an area image where the object in the multispectral image changes.

[0062] In the embodiment of the present invention, considering that when a transmission line fails, its typical components will undergo some morphological changes, differential image technology is used to detect areas where object morphology changes occur in the target transmission line during the shooting time. These change areas correspond to dynamic target areas, so there may be loose components or foreign objects invading the dynamic target area.

[0063] Step 150: Perform noise reduction analysis on the dynamic area image according to the geographic information and meteorological information to obtain an abnormal area image of the component in the target transmission line.

[0064] Specifically, meteorological information is obtained when shooting the multispectral sequence, and the meteorological factors are analyzed based on the meteorological information to determine the dynamic effects on objects in the area shot by the multispectral camera. For example, when the wind speed is too high, the objects in the area shot undergo significant morphological changes. Since the dynamic changes caused by meteorological factors are not abnormal changes caused by power transmission line faults, they may lead to misjudgment of fault detection. In order to improve the accuracy of the detection results, in the embodiment of the present invention, the dynamic changes caused by meteorological factors on the target power transmission line and objects in its surrounding environment are excluded to reduce the noise caused by meteorological factors on the dynamic area image.

[0065] Furthermore, since the camera does not only shoot the transmission line area when photographing the target transmission line, the surrounding environment of the target transmission line may also exist in the obtained multispectral image. In order to avoid the noise caused by changes in the surrounding environment to abnormal phenomena in the transmission line, the position of the target transmission line on the multispectral image can also be determined through geographic information, and other images except the target transmission line in the dynamic area image are excluded, and only the image containing the dynamic area of ​​the target transmission line is retained. The processed dynamic area image only contains the abnormal phenomena existing in the target transmission line, so the image is an image of the abnormal area of ​​the component in the target transmission line.

[0066] The embodiment of the present invention performs noise reduction analysis on the dynamic area image by eliminating the influence of geographical factors and meteorological factors on the photographed area when shooting the multispectral image, and obtains an image of the abnormal area of ​​the component that only contains the target transmission line, so as to reduce the error caused by atypical component failures to the identification result.

[0067] Step 170, collect abnormal features in the image of the abnormal region of the component, and input the abnormal features into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm, where the recognition result is that the component in the abnormal region of the component is normal or the component in the abnormal region of the component is faulty.

[0068] In the embodiment of the present invention, abnormal features are extracted from the abnormal region image according to the characteristics of the component and the prominent features when the component fails, so as to distinguish the faulty components in the abnormal region image based on the content of the obvious abnormal features.

[0069] The preset pattern recognition algorithm may be an applied machine learning or deep learning algorithm, such as a support vector machine, a convolutional neural network, etc. The pattern recognition algorithm is a classification model of abnormal features to distinguish between actual component abnormal areas and normal areas in the component abnormal area image.

[0070] The present invention captures information of different bands by collecting multispectral image sequences of transmission lines, so that image data can be effectively obtained even under poor lighting conditions, thereby improving the accuracy of detection results. In addition, by combining meteorological information and geographic information to denoise dynamic area images, the interference of geographical and meteorological factors on detection results can be reduced, making the detection results more accurate.

[0071] In one embodiment of the present invention, step 130, performing image analysis on the multispectral image sequence based on the differential image technology to obtain the dynamic area image where the target transmission line exists, specifically includes:

[0072] Step 310, select any two adjacent multispectral images in the multispectral image sequence, and use the multispectral images of the two adjacent frames as the target multispectral images.

[0073] Step 320, preprocessing the target multispectral image to obtain a denoised target multispectral image, wherein the preprocessing at least includes normalization and / or brightness correction.

[0074] In the embodiment of the present invention, two adjacent frames of multispectral images are selected in the multispectral image sequence as objects of differential image processing, ie, target multispectral images.

[0075] The two frames of target multispectral images are preprocessed to make the sizes of the two frames of target multispectral images consistent and correct the differences in brightness, thereby reducing the impact of changes in factors such as lighting or shooting angles.

[0076] Step 330, calculate the pixel difference between the target multispectral images after noise reduction to obtain a differential image.

[0077] Specifically, the points belonging to the same geographic information in two adjacent frames of target multispectral images are compared, and the pixel-level difference values ​​between pixels with the same geographic information are calculated to form a differential image, which highlights the pixel areas that have changed in the two frames of images. In the differential image, the pixel areas that have changed will be highlighted to help identify potential dynamic target areas.

[0078] Step 340, denoise the pixel area in the differential image to obtain a dynamic area image where the target transmission line exists.

[0079] Specifically, a noise reduction process is performed on smaller changes in the differential image to remove the influence of noise and retain the dynamic region image with significant changes.

[0080] In one embodiment of the present invention, Step 340, denoising the pixel area in the differential image to obtain a dynamic area image where the target power transmission line exists, specifically includes:

[0081] Step 341, denoise the differential image based on threshold processing technology to obtain the initial dynamic area image after denoising.

[0082] Specifically, the threshold processing technology is used to determine the optimal pixel value range, retain the pixel points within the optimal pixel value range, delete the pixel points that are not within the optimal pixel value range, obtain the area affected by the filtered noise, and only retain those areas with significant changes as potential initial dynamic target areas.

[0083] Step 342, perform morphological operations on the initial dynamic area image to obtain the target dynamic area image.

[0084] Specifically, morphological operations such as dilation and erosion can be used to remove small bright spot noise and small dark spot noise.

[0085] After the initial dynamic region image is subjected to morphological operation, the connected domain in the initial dynamic region image is processed to eliminate isolated noise points and merge adjacent connected domains to obtain the processed target dynamic region image. The connected domain in the differential image is processed by morphological operation to remove isolated noise points and merge adjacent connected domains, making the region in the target dynamic region image more clear and concentrated.

[0086] In one embodiment of the present invention, step 150, performing noise reduction analysis on the dynamic area image according to the geographic information and the meteorological information to obtain an abnormal area image of the component in the target transmission line, specifically includes:

[0087] Step 510, searching for target meteorological data of the area where the dynamic area image is located from the geographic information and meteorological information.

[0088] Specifically, the precise location information of the target transmission line, including the line direction, tower location and height, etc., is retrieved from the Geographic Information System (GIS) database, and this information is matched with the geographic coordinates in the dynamic area image to ensure that every pixel in the image can correspond to the actual geographic location.

[0089] Access to real-time meteorological services to obtain meteorological data at the time of multispectral image sequence shooting, including wind speed, temperature, humidity, visibility, etc., helps to understand the weather conditions at the time of shooting, and is used for subsequent analysis to consider the impact of weather on changes in objects within the target transmission line.

[0090] Step 520, analyzing the dynamic impact of the target meteorological data on the non-power facilities in the dynamic area image, and obtaining the first false positive dynamic area in the dynamic area image, where the false positive dynamic area is the area where the non-power facilities in the dynamic area image change.

[0091] Specifically, the impact of meteorological factors on the potential dynamic area image is analyzed based on the meteorological information when the image is taken. For example, the image changes caused by the swaying of trees or the movement of other non-power facilities due to wind. By combining meteorological data with the dynamic target area image, the impact of meteorological factors on the changing area in the image is analyzed to exclude the first false positive dynamic area caused by these natural phenomena.

[0092] Step 530: exclude the first false positive dynamic region in the dynamic region image to obtain a standard dynamic region image.

[0093] Step 540: Determine the actual line area of ​​the target transmission line in the standard dynamic area image according to the geographic information of the target transmission line.

[0094] Specifically, the dynamic area outside the target transmission line in the standard dynamic area image is excluded. Combined with the accurate geographic information of the target transmission line in the GIS data, the actual position relationship between the standard dynamic area image and the target transmission line is compared, and the dynamic target outside the target transmission line in the standard dynamic area image is excluded.

[0095] Step 550: exclude other areas in the standard dynamic area image except the actual line area, and obtain an abnormal area image of the component of the target transmission line.

[0096] In one embodiment of the present invention, the historical fault data and maintenance records of the target transmission line in the GIS can also be used to evaluate whether the target transmission line has ever experienced similar conditions affected by weather factors, and further analysis can be performed to determine the abnormal area of ​​the component to increase the credibility of the abnormality detection.

[0097] The present invention combines geographic information with meteorological information to exclude false positive areas caused by weather changes and components outside the transmission line, thereby ensuring the authenticity of the detection results. In addition, through the analysis of historical fault data and maintenance records, the credibility of anomaly detection is improved, making the location of faulty components more accurate.

[0098] In one embodiment of the present invention, the method further includes:

[0099] Step 161: retrieve the positioning information of the component abnormal area image from the geographic information of the target transmission line, and determine the component abnormal area of ​​the target transmission line according to the positioning information.

[0100] Specifically, the specific location of the component abnormal area image is found through the geographic information of the target transmission line.

[0101] Step 162: deploy a structured lighting device at the abnormal component area of ​​the target transmission line, and control the structured lighting device to project structured light to the abnormal component area according to a preset lighting mode.

[0102] In order to observe the abnormal area of ​​the component more clearly, structured lighting technology is used to deploy structured lighting equipment in the abnormal area of ​​the component, such as LED array light source or other programmable light source, to provide a preset lighting mode to illuminate the abnormal area of ​​the component, such as stripe light, spot light or other coded patterns. This lighting method can enhance the texture contrast of the surface of the object, making the details of the abnormal area more obvious and facilitating subsequent analysis.

[0103] In the embodiment of the present invention, step 170, collecting abnormal features in the image of the abnormal region of the component, and inputting the abnormal features into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm, specifically includes:

[0104] Step 710, collect structured light images of abnormal areas of the components.

[0105] Specifically, a high-resolution camera or a dedicated texture capture camera installed on the same drone is used to capture a structured light image of the abnormal area of ​​the component under the illumination of a structured lighting device.

[0106] Step 710, extracting abnormal features from the structured light image, and inputting the abnormal features into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm.

[0107] Specifically, the abnormal features in the structured light image are extracted, the abnormal features are classified based on a preset pattern recognition algorithm, and the abnormal parts in the structured light image are determined.

[0108] In one embodiment of the present invention, Step 710, extracting abnormal features from the structured light image, and inputting the abnormal features into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm, specifically includes:

[0109] Step 711, decode the structured light image based on a preset image processing technology to obtain depth information of the structured light image.

[0110] Specifically, the structured light image is processed by a preset image processing technology, such as a phase shift method or a grayscale center method, to decode the structured light image to obtain the depth information of the surface of the abnormal area of ​​the component, and the contrast of the image is enhanced based on the depth information of the surface of the abnormal area to facilitate subsequent image analysis.

[0111] Step 720, enhancing the texture contrast in the structured light image according to the depth information of the structured light image to obtain an enhanced texture image.

[0112] Step 730, using a preset feature extraction algorithm to extract texture features in the texture image, inputting the texture features into a pattern recognition algorithm, and obtaining a recognition result output by the pattern recognition algorithm.

[0113] Specifically, the texture features in the texture image are extracted using a preset feature extraction algorithm, which may be a scale-invariant feature transform (IFT) and a speeded-up robust feature (SURF), etc. High-contrast key feature points in the abnormal area of ​​the component, such as corners, edges, spots, etc., are extracted, and texture features are calculated based on the feature points, such as statistical characteristics, such as mean and variance; frequency characteristics, such as Fourier transform; spatial distribution characteristics, such as histogram of oriented gradients (HOG), etc., so as to input the texture features with high contrast into the pattern recognition algorithm for fault diagnosis.

[0114] The embodiments of the present invention improve the visibility of abnormal parts in the image through structured lighting technology, reduce the detection errors caused by poor lighting conditions, and ensure the reliability of the detection results.

[0115] In one embodiment of the present invention, high-contrast texture features can be combined with the positioning information of previously identified abnormal areas of components to generate a detailed problem location diagram, providing intuitive visual support for subsequent fault confirmation and repair, allowing the maintenance team to locate the problem more accurately.

[0116] In one embodiment of the present invention, a pattern recognition algorithm is used to distinguish between normal components and faulty components in the component abnormal region image, and to locate the specific position of the faulty component, which can be achieved in the following manner:

[0117] First, the texture features are input into a pre-trained pattern recognition model for classification. These models can be supervised learning models such as support vector machines (SVM), decision trees, random forests, or deep learning models such as convolutional neural networks (CNN). These models have learned the decision boundaries to distinguish between normal and faulty parts during the training phase.

[0118] For example, if a support vector machine (SVM) is used, the decision function of the model can be expressed as:

[0119]

[0120] Where x is the feature vector of the input texture feature, y i is the label of the training sample, which is normal or faulty, α i is the Lagrange multiplier, K(·,·) is the kernel function, and b is the bias term.

[0121] After classifying the texture features, it is necessary to further locate the position of the faulty component. This can be achieved by finding the bounding box of the faulty area in the structured light image. Object detection technology is used to output one or more bounding boxes, each of which contains a location information and a category label. The location information contains the upper left corner coordinates, width and height data, and the category label is normal or faulty.

[0122] Finally, the position of the faulty component in the structured light image is mapped back to the coordinate system in the real world, and the faulty component is located using the known scale of the image and the real world or through GPS data and the attitude information of the drone.

[0123] In one embodiment of the present invention, the method also includes: when the identification result is a component failure within a component abnormality area, retrieving the positioning data of the faulty component in the geographic information of the target transmission line; generating a maintenance instruction for the faulty component based on the positioning data of the faulty component, and outputting the maintenance instruction so as to repair the faulty component based on the maintenance instruction.

[0124] Specifically, after detecting a faulty component, the location data of the faulty component is integrated, including the geographic coordinates and the precise location description relative to the target transmission line, so that the maintenance team can quickly locate the problem. According to the type of faulty component and its location data, a maintenance instruction is generated, which includes maintenance suggestions, required tools and spare parts lists. The maintenance instruction provides the maintenance team with specific operating guidelines to ensure the smooth progress of maintenance work.

[0125] In a feasible embodiment of the present invention, the maintenance instruction also includes photos or video materials of the faulty parts to facilitate the maintenance team to intuitively understand the on-site conditions. These multimedia materials help maintenance personnel to more accurately judge the fault conditions and prepare corresponding maintenance measures.

[0126] Using the wireless communication module on the drone or in the fixed equipment, the maintenance instructions are sent to the remote server or directly pushed to the mobile terminal device of the maintenance team to ensure that the maintenance information can quickly reach the maintenance personnel so that the maintenance team can maintain the faulty parts in time. In order to adapt to various working environments, the wireless communication module supports multiple communication protocols, which can ensure the effective transmission of data whether in cities or remote areas. This feature ensures the reliability and coverage of information transmission.

[0127] In one embodiment of the present invention, when the maintenance team receives a maintenance instruction, the sending time and receiving status of the maintenance instruction are recorded, which provides a basis for subsequent tracking and inquiry, helps manage the maintenance progress and ensures the transparency of the maintenance work.

[0128] The present invention also provides a detection device for a faulty component of a power transmission line, which can be referred to Figure 2 , Figure 2 2 is a structural block diagram of a device for detecting a faulty component of a power transmission line according to an embodiment of the present invention. The device includes a data acquisition unit 201 , an image analysis unit 202 and an abnormality determination unit 203 .

[0129] The data acquisition unit 201 is used to acquire a multispectral image sequence of a target power transmission line, as well as geographic information and meteorological information of the target power transmission line.

[0130] The image analysis unit 202 is used to perform image analysis on the multispectral image sequence based on the differential image technology to obtain a dynamic area image where the target transmission line exists. The dynamic area image is an area image where the object in the multispectral image changes.

[0131] The dynamic area image is subjected to noise reduction analysis based on geographic information and meteorological information to obtain the abnormal area image of the components in the target transmission line.

[0132] The abnormality judgment unit 203 is used to collect abnormal features in the image of the component abnormal area, and input the abnormal features into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm, where the recognition result is that the component in the component abnormal area is normal or the component in the component abnormal area is faulty.

[0133] The detection device for faulty components of a transmission line proposed in an embodiment of the present invention captures information of different bands by collecting a multispectral image sequence of the transmission line, so that image data can be effectively obtained even under poor lighting conditions, thereby improving the accuracy of the detection results; secondly, by combining meteorological information and geographic information to denoise the dynamic area image, the interference of geographical and meteorological factors on the detection results can be reduced, making the detection results more accurate.

[0134] Figure 3 FIG. 1 shows an internal structure diagram of a computer device in one embodiment of the present invention. The computer device may be a terminal or a system. Figure 3 As shown, the computer device includes a processor, a memory and a network interface connected via a system bus. The memory includes a non-volatile storage medium and an internal memory. The non-volatile storage medium of the computer device stores an operating system and may also store a computer program. When the computer program is executed by the processor, the processor may implement each step in the above method embodiment. The internal memory may also store a computer program. When the computer program is executed by the processor, the processor may implement each step in the above method embodiment. Those skilled in the art will understand that Figure 3 The structure shown in the figure is only a block diagram of a part of the structure related to the solution of the present application, and does not constitute a limitation on the computer device to which the solution of the present application is applied. The specific computer device may include more or fewer components than those shown in the figure, or combine certain components, or have a different arrangement of components.

[0135] In one embodiment, a computer device is provided, including a memory and a processor, wherein the memory stores a computer program, and when the computer program is executed by the processor, the processor executes each step in the above method embodiment.

[0136] In one embodiment, a computer-readable storage medium is proposed, which stores a computer program. When the computer program is executed by a processor, the processor executes each step in the above method embodiment. A person of ordinary skill in the art can understand that all or part of the processes in the above embodiment method can be completed by instructing the relevant hardware through a computer program. The program can be stored in a non-volatile computer-readable storage medium, and the program can include the processes of the embodiments of the above methods when executed. Among them, any reference to memory, storage, database or other media used in the embodiments provided in this application may include non-volatile and / or volatile memory. Non-volatile memory may include read-only memory (ROM), programmable ROM (PROM), electrically programmable ROM (EPROM), electrically erasable programmable ROM (EEPROM) or flash memory. Volatile memory may include random access memory (RAM) or external cache memory. By way of illustration and not limitation, RAM is available in many forms, such as static RAM (SRAM), dynamic RAM (DRAM), synchronous DRAM (SDRAM), double data rate SDRAM (DDRSDRAM), enhanced SDRAM (ESDRAM), synchronous link (Synchlink) DRAM (SLDRAM), memory bus (Rambus) direct RAM (RDRAM), direct memory bus dynamic RAM (DRDRAM), and memory bus dynamic RAM (RDRAM), etc.

[0137] The technical features of the above embodiments may be combined arbitrarily. To make the description concise, not all possible combinations of the technical features in the above embodiments are described. However, as long as there is no contradiction in the combination of these technical features, they should be considered to be within the scope of this specification.

[0138] The above embodiments only express several implementation methods of the present application, and the descriptions thereof are relatively specific and detailed, but they cannot be understood as limiting the scope of the present application. It should be pointed out that, for a person of ordinary skill in the art, several modifications and improvements can be made without departing from the concept of the present application, and these all belong to the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the attached claims.

Claims

1. A method for detecting faulty components of a power transmission line, characterized in that: The method comprises: Collecting a multispectral image sequence of a target transmission line, as well as geographic information and meteorological information of the target transmission line; Performing image analysis on the multispectral image sequence based on differential image technology to obtain a dynamic area image where the target transmission line exists, wherein the dynamic area image is an area image where an object in the multispectral image changes; Performing noise reduction analysis on the dynamic area image according to the geographic information and meteorological information to obtain an abnormal area image of a component in the target transmission line; Abnormal features in the image of the abnormal area of ​​the component are collected, and the abnormal features are input into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm, wherein the recognition result is that the component in the abnormal area of ​​the component is normal or the component in the abnormal area of ​​the component is faulty.

2. The method according to claim 1, characterized in that The method further comprises: Retrieving the positioning information of the component abnormal area image from the geographic information of the target transmission line, and determining the component abnormal area of ​​the target transmission line according to the positioning information; Deploy a structured lighting device at the abnormal region of the component of the target power transmission line, and control the structured lighting device to project structured light to the abnormal region of the component according to a preset lighting mode; The collecting of abnormal features in the image of the abnormal region of the component and inputting the abnormal features into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm specifically includes: Acquiring a structured light image of an abnormal region of the component; Abnormal features in the structured light image are extracted, and the abnormal features are input into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm.

3. The method according to claim 2, characterized in that The extracting of abnormal features from the structured light image and inputting the abnormal features into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm specifically includes: Decoding the structured light image based on a preset image processing technology to obtain depth information of the structured light image; enhancing the texture contrast in the structured light image according to the depth information of the structured light image to obtain an enhanced texture image; The texture features in the texture image are extracted using a preset feature extraction algorithm, and the texture features are input into the pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm.

4. The method according to claim 1, characterized in that The performing of image analysis on the multispectral image sequence based on the differential image technology to obtain the dynamic area image where the target transmission line exists specifically includes: Selecting any two adjacent multispectral images in the multispectral image sequence, and using the multispectral images of the two adjacent frames as target multispectral images; Preprocessing the target multispectral image to obtain a denoised target multispectral image, wherein the preprocessing includes at least normalization and / or brightness correction; Calculating the pixel difference between the denoised target multispectral images to obtain a differential image; The pixel area in the differential image is denoised to obtain a dynamic area image where the target power transmission line exists.

5. The method according to claim 4, characterized in that The step of removing noise from the pixel area in the differential image to obtain the dynamic area image where the target power transmission line exists specifically includes: De-noising the differential image based on a threshold processing technique to obtain a de-noised initial dynamic region image; A morphological operation is performed on the initial dynamic region image to obtain a target dynamic region image.

6. The method according to claim 1, characterized in that The performing noise reduction analysis on the dynamic area image according to the geographic information and the meteorological information to obtain the abnormal area image of the component in the target transmission line specifically includes: Searching for target meteorological data of the area where the dynamic area image is located from the geographic information and the meteorological information; Analyze the dynamic impact of the target meteorological data on the non-power facilities in the dynamic area image to obtain a first false positive dynamic area in the dynamic area image, where the false positive dynamic area is an area in the dynamic area image where the non-power facilities change; Excluding the first false positive dynamic area in the dynamic area image to obtain a standard dynamic area image; Determining an actual line area of ​​the target transmission line in the standard dynamic area image according to the geographic information of the target transmission line; The other regions in the standard dynamic region image except the actual line region are excluded to obtain the component abnormal region image of the target power transmission line.

7. The method according to claim 1, characterized in that The method further comprises: When the identification result is a component failure within the component abnormality area, retrieving the location data of the faulty component from the geographic information of the target transmission line; A maintenance instruction for the faulty component is generated according to the location data of the faulty component, and the maintenance instruction is outputted so as to perform maintenance on the faulty component based on the maintenance instruction.

8. A device for detecting faulty components of a power transmission line, characterized in that: The device comprises a data acquisition unit, an image analysis unit and an abnormality determination unit; The data acquisition unit is used to acquire a multispectral image sequence of a target power transmission line, as well as geographic information and meteorological information of the target power transmission line; The image analysis unit is used to perform image analysis on the multispectral image sequence based on differential image technology to obtain a dynamic area image where the target transmission line exists, wherein the dynamic area image is an area image where an object in the multispectral image changes; Performing noise reduction analysis on the dynamic area image according to the geographic information and meteorological information to obtain an abnormal area image of a component in the target transmission line; The abnormality judgment unit is used to collect abnormal features in the image of the abnormal area of ​​the component, and input the abnormal features into a preset pattern recognition algorithm to obtain a recognition result output by the pattern recognition algorithm, wherein the recognition result is that the component in the abnormal area of ​​the component is normal or the component in the abnormal area of ​​the component is faulty.

9. A computer-readable storage medium storing a computer program, characterized in that: When the computer program is executed by a processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 7.

10. A computer device comprising a memory and a processor, characterized in that: The memory stores a computer program, and when the computer program is executed by the processor, the processor is caused to perform the steps of the method according to any one of claims 1 to 7.

Images