A power transmission line multi-bundle conductor detection method, device, equipment and medium

Abstract

The application discloses a power transmission line multi-bundle conductor detection method, device, equipment and medium. The method comprises the following steps: determining a region of interest of the multi-bundle conductor in the aerial image according to the motion state of the unmanned aerial vehicle, and determining a brightness threshold corresponding to the aerial image according to the pixel brightness in the region of interest; determining a plurality of conductor candidate pixels in the region of interest according to the brightness threshold, and performing clustering processing on the conductor candidate pixels according to the positions of the conductor candidate pixels to generate conductor clusters; if the number of conductor clusters generated after the clustering processing exceeds a first value, identifying effective conductor clusters in the conductor clusters according to the number of pixels in each conductor cluster; extracting multi-dimensional feature parameters of the effective conductor clusters, and generating a visual label in the aerial image according to the multi-dimensional feature parameters of the effective conductor clusters. The above technical solution can avoid clustering merging or missed detection and other phenomena, adapt to brightness differences in different aerial scenes, and effectively improve the detection accuracy.

Description

Technical Field

[0001] This invention relates to the fields of computer vision and intelligent detection technology for power transmission lines, and in particular to a method, apparatus, equipment and medium for detecting multi-split conductors in power transmission lines. Background Technology

[0002] In high-voltage transmission lines, multi-split conductors are often used to improve transmission efficiency. Multi-split conductors are parallel conductor structures in which a single phase conductor is split into two or more sub-conductors of the same type, and the spacing is fixed by spacers and arranged symmetrically.

[0003] In existing technologies, conductor detection of power transmission lines can be performed using detection methods based on traditional image processing or deep learning.

[0004] However, detection methods based on traditional image processing have problems such as weak resistance to background interference, sensitivity to changes in lighting, and inability to adapt to densely distributed multi-splitting wires. Although deep learning-based methods have higher accuracy, their model structures are complex and computationally expensive, making them difficult to deploy in UAV embedded inspection equipment. Furthermore, multi-splitting wires are characterized by dense distribution and small spacing, and existing algorithms lack targeted clustering and separation strategies, which can easily lead to wire clustering merging or missed detections, resulting in low detection accuracy. Summary of the Invention

[0005] This invention provides a method, apparatus, equipment, and medium for detecting multi-split conductors in power transmission lines. It can avoid clustering or missed detection phenomena in scenarios such as changing lighting and complex backgrounds, adapt to brightness differences in different aerial photography scenarios, and effectively improve detection accuracy.

[0006] According to one aspect of the present invention, a method for detecting multi-split conductors in transmission lines is provided, comprising: Based on the movement state of the UAV, the region of interest for the multi-split conductor is determined in the aerial image, and the brightness threshold corresponding to the aerial image is determined based on the pixel brightness within the region of interest. Based on the brightness threshold, multiple candidate pixels for conductors are determined within the region of interest, and the candidate pixels for conductors are clustered according to the position of each candidate pixel to generate a conductor cluster. If the number of wire clusters generated after clustering exceeds the first value, then the valid wire clusters are identified in each wire cluster based on the number of pixels in each wire cluster. Extract multi-dimensional feature parameters of effective conductor clusters, and generate visual annotations in aerial images based on these parameters.

[0007] According to another aspect of the present invention, a device for detecting multi-split conductors in transmission lines is provided, comprising: The brightness threshold determination module is used to determine the region of interest of the multi-split conductor in the aerial image based on the motion state of the UAV, and to determine the brightness threshold corresponding to the aerial image based on the pixel brightness in the region of interest. The conductor cluster generation module is used to determine multiple conductor candidate pixels in the region of interest according to the brightness threshold, and to perform clustering processing on the conductor candidate pixels according to the position of each conductor candidate pixel to generate a conductor cluster. The effective conductor cluster identification module is used to identify effective conductor clusters in each conductor cluster based on the number of pixels in each conductor cluster if the number of conductor clusters generated after clustering exceeds a first value. The visualization processing module is used to extract multi-dimensional feature parameters of effective conductor clusters and generate visual annotations in aerial images based on these parameters.

[0008] According to another aspect of the present invention, an electronic device is provided, the electronic device comprising: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, which enables the at least one processor to perform the transmission line multi-split conductor detection method according to any embodiment of the present invention.

[0009] According to another aspect of the present invention, a computer-readable storage medium is provided, the computer-readable storage medium storing computer instructions for causing a processor to execute and implement the transmission line multi-split conductor detection method according to any embodiment of the present invention.

[0010] The technical solution of this invention involves determining the region of interest (ROI) for a multi-splitting duct in an aerial image based on the motion state of the UAV, determining a brightness threshold corresponding to the aerial image based on the pixel brightness within the ROI, identifying multiple candidate duct pixels within the ROI based on the brightness threshold, and clustering these candidate pixels according to their positions to generate duct clusters. If the number of duct clusters generated after clustering exceeds a first value, valid duct clusters are identified within each cluster based on the number of pixels within each cluster. Multi-dimensional feature parameters of the valid duct clusters are then extracted, and based on... The multi-dimensional feature parameters of effective conductor clusters are used to generate visual annotations in aerial images. This method can dynamically determine the brightness threshold for the currently processed aerial image, solving the problem that fixed thresholds cannot adapt to different lighting conditions. It also solves the clustering and merging problem of multi-split conductors due to small spacing, ensuring accurate separation of multi-split conductors. For scenarios with too many conductor clusters after clustering, the number of pixels is used as the core screening criterion to quickly remove pseudo-clusters, avoiding interference from pseudo-clusters in the detection results and effectively improving the detection accuracy of conductor clusters. At the same time, it takes into account the real-time performance and robustness of the detection, and adapts to the computing power requirements of UAV embedded devices.

[0011] It should be understood that the description in this section is not intended to identify key or essential features of the embodiments of the present invention, nor is it intended to limit the scope of the invention. Other features of the invention will become readily apparent from the following description. Attached Figure Description

[0012] To more clearly illustrate the technical solutions in the embodiments of the present invention, the accompanying drawings used in the description of the embodiments will be briefly introduced below. Obviously, the accompanying drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0013] Figure 1 This is a flowchart of a method for detecting multi-split conductors in transmission lines according to Embodiment 1 of the present invention; Figure 2 This is a flowchart of another method for detecting multi-split conductors in transmission lines according to Embodiment 2 of the present invention; Figure 3 This is a schematic diagram of the structure of a multi-split conductor detection device for transmission lines according to Embodiment 3 of the present invention; Figure 4 This is a schematic diagram of the structure of an electronic device that implements the method for detecting multi-split conductors in power transmission lines according to an embodiment of the present invention. Detailed Implementation

[0014] To enable those skilled in the art to better understand the present invention, the technical solutions of the present invention will be clearly and completely described below with reference to the accompanying drawings of the embodiments of the present invention. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort should fall within the scope of protection of the present invention.

[0015] It should be noted that the terms "first," "second," etc., in the specification, claims, and accompanying drawings of this invention are used to distinguish similar objects and are not necessarily used to describe a specific order or sequence. It should be understood that such terms can be used interchangeably where appropriate so that embodiments of the invention described herein can be implemented in orders other than those illustrated or described herein. Furthermore, the terms "comprising" and "having," and any variations thereof, are intended to cover a non-exclusive inclusion; for example, a process, method, system, product, or apparatus that comprises a series of steps or units is not necessarily limited to those steps or units explicitly listed, but may include other steps or units not explicitly listed or inherent to such processes, methods, products, or apparatus.

[0016] Example 1 Figure 1 This is a flowchart of a method for detecting multi-branched conductors in transmission lines according to Embodiment 1 of the present invention. This embodiment is applicable to the automated identification and parameter analysis of multi-branched conductors in aerial images of transmission lines. The method can be executed by a multi-branched conductor detection device, which can be implemented in hardware and / or software and is generally configured in a computer or processor with data processing capabilities. Figure 1 As shown, the method includes: S110. Based on the motion state of the UAV, determine the region of interest for the multi-split duct in the aerial image, and determine the brightness threshold corresponding to the aerial image based on the pixel brightness within the region of interest.

[0017] Optionally, the region of interest determined in the aerial image is the area where the conductors are likely to be distributed and the background is relatively clean. In general, most of the complex background can be masked in the region of interest, thereby reducing the amount of invalid pixels to be processed and reducing the computational complexity of the algorithm.

[0018] Optionally, when the aerial image is an RGB color image, it can be converted into a grayscale image. The single pixel value in the grayscale image directly represents the pixel brightness. The brightness threshold is a critical value used to distinguish the brightness of wire pixels from background pixels in the aerial image. Pixels with brightness higher than this value are determined to be wire candidate pixels, and vice versa. The brightness threshold in this invention can be dynamically constructed according to the currently processed aerial image, thereby adapting to different UAV flight environments.

[0019] Specifically, based on the motion state of the UAV, the region of interest for the multi-split duct is determined in the aerial image, and a brightness threshold corresponding to the aerial image is determined based on the pixel brightness within the region of interest, including: Based on the streaming video captured by the UAV, the flight direction of the UAV relative to the video frame is determined, and based on the flight direction and the typical distribution position of the conductors corresponding to the flight direction, the region of interest of the multi-split conductors is determined in the aerial image; wherein, the aerial image is a single frame video image in the streaming video. Based on the pixel brightness within the region of interest and the lower limit constraint for brightness, a first lower limit for brightness is determined, and based on the first lower limit for brightness, each effective pixel within the region of interest whose brightness is greater than the first lower limit for brightness is determined. Based on the brightness of each effective pixel, a brightness threshold corresponding to the aerial image is determined.

[0020] Optionally, the streaming video acquired by the UAV consists of continuous single-frame aerial images. By analyzing the inter-frame changes of the streaming video, the flight direction of the UAV relative to the video frame is determined, such as horizontal left flight or horizontal right flight. Based on this flight direction, the typical distribution position of the conductor is matched with a pre-set parameter, and the region of interest of the multi-split conductor is delineated in the current aerial image.

[0021] Optionally, the region of interest can be a specified rectangular area in the aerial image. The region of interest can be determined in the aerial image by defining the row and column ranges by the typical distribution location of the traverse.

[0022] Optionally, since the multi-splitter conductors are supported and fixed by the transmission towers, the flight direction of the drone relative to the video frame can also represent the drone's flight direction between the transmission towers. Technicians can pre-configure the typical distribution position of the conductors corresponding to each flight direction. For example, during the process of flying from the left transmission tower to the right transmission tower, the left background of most image frames in the streaming video may contain the image of the left transmission tower, which is prone to interference. Therefore, when the drone flies in this direction, the typical distribution position of the conductors can be set in a specified area on the right side of the image. This is only an example.

[0023] Optionally, the first brightness lower limit is a brightness threshold determined based on the pixel brightness characteristics of the region of interest and the brightness lower limit constraint. It is used to remove low-brightness background pixels in the region of interest to avoid them interfering with the calculation of the brightness threshold. Pixels in the region of interest with brightness higher than the first brightness lower limit can be identified as valid pixels. Then, the dynamic brightness threshold is calculated based on the valid pixels, thereby eliminating interference from invalid pixels such as pure black backgrounds and low-brightness noise.

[0024] Optionally, the brightness lower limit constraint can be a fixed brightness lower limit value, or it can be a value dynamically calculated based on the pixel brightness in the region of interest and a fixed offset. There is no limitation here.

[0025] Optionally, since aerial images often feature high brightness of conductors, the brightness threshold can be determined by calculating the 90th percentile of brightness. Specifically, the brightness of all valid pixels can be sorted from darkest to brightest, and the brightness values ​​in the top 90% can be used as the 90th percentile of brightness. Alternatively, the brightness threshold can also be determined by calculating the mean, variance, or other forms of brightness. The specific calculation method can be determined by technicians based on experience.

[0026] Optionally, the brightness threshold can be calculated by combining the 90th percentile of the brightness of the effective pixels and the standard deviation. Specifically, it can be calculated using the following formula: BRIGHT=P90-k×φ, where P90 is the 90th percentile of the brightness of the effective pixels, k is a preset adaptive coefficient, φ is the standard deviation of the effective pixels, and BRIGHT is the brightness threshold. This is only an example for illustration.

[0027] S120. Based on the brightness threshold, determine multiple candidate pixels for conductors within the region of interest, and cluster the candidate pixels for conductors according to their positions to generate conductor clusters.

[0028] Optionally, based on a brightness threshold, all pixels within the region of interest can be binarized, retaining pixels with brightness above the brightness threshold as candidate wire pixels, and setting pixels with brightness below the brightness threshold as background, thereby achieving rapid separation between candidate wire pixels and background pixels, with the brightness of candidate wire pixels conforming to wire characteristics.

[0029] Optionally, a conductor cluster is a pixel cluster formed by grouping adjacent conductor candidate pixels into the same set. One conductor cluster corresponds to a multi-splitting conductor in the aerial image. Specifically, one-dimensional greedy clustering can be performed based on the positional features of the conductor candidate pixels in the image to group adjacent pixels into the same cluster and generate conductor clusters. The clustering interval of the one-dimensional greedy clustering can be set to a small value, such as 2 pixels, so as to avoid the clustering merging problem of multi-splitting conductors due to small intervals through small-interval clustering design.

[0030] Specifically, based on the brightness threshold, multiple candidate pixels for conductors are determined within the region of interest, and based on the position of each candidate pixel, the candidate pixels are clustered to generate conductor clusters, including: Based on the brightness threshold, each pixel in the region of interest is binarized and segmented, and the pixels retained after binarization are determined as candidate pixels for the conductor. Based on the vertical coordinate values ​​of each candidate pixel in the aerial image, the candidate pixels are sorted, and based on the sorting results and the pre-set clustering interval, adjacent candidate pixels with a vertical distance less than the clustering interval are divided into the same candidate pixel cluster.

[0031] Optionally, if the brightness value of a pixel is higher than the brightness threshold, the pixel is retained and set to white; if the brightness value of a pixel is less than or equal to the brightness threshold, the pixel is discarded and set to black. After all pixels in the region of interest are divided in the above manner, the aerial image is converted into a black and white image. The white area represents the initially divided guide line area, and the pixels in the white area are candidate pixels for the guide line. The black area represents the initially divided background area.

[0032] Optionally, binarization segmentation refers to converting the pixel values ​​of the grayscale aerial image into 0 or 255, where 0 is used to identify black background areas and 255 is used to identify white target areas. The clustering distance is the core parameter for determining whether a candidate pixel of a conductor is a critical value of the distance between adjacent pixels.

[0033] Optionally, the candidate pixels of the conductor can be sorted in ascending order according to their ordinate values; a greedy clustering algorithm is used to traverse the candidate pixels of the conductor from beginning to end according to the sorting results, and adjacent candidate pixels of the conductor whose ordinate distance is less than the clustering interval are divided into the same conductor cluster.

[0034] Optionally, the purpose of using a greedy clustering algorithm is to traverse each candidate pixel of a wire from beginning to end in the sorted candidate pixels of the wire, and only look at the current nearest point. As long as the distance from the previous pixel does not exceed the clustering interval, it is merged into the same wire. If it exceeds the clustering interval, it is treated as another wire. The greedy clustering algorithm does not perform complex calculations and is extremely fast.

[0035] S130. If the number of wire clusters generated after clustering exceeds the first value, then the valid wire clusters are identified in each wire cluster based on the number of pixels in each wire cluster.

[0036] Optionally, since a multi-split duct is composed of multiple sub-ducts of the same type, a first value can be set according to the number of sub-ducts. For example, when the number of sub-ducts in a multi-split duct is 6, the first value can be set to 4. If the number of duct clusters does not exceed the first value, the currently acquired aerial image may be of poor quality, and the aerial image may not cover most of the sub-ducts. In this case, the aerial image can be discarded. At the same time, if the number of duct clusters is too small, it will also reduce the accuracy of pseudo-cluster segmentation. For example, if the number of duct clusters is 1, without comparison between multiple duct clusters, it is impossible to determine whether the duct cluster is a pseudo-cluster.

[0037] Optionally, an effective conductor cluster refers to a conductor cluster that has been filtered according to rules and after removing false clusters, and whose pixel count, morphological features, etc., all conform to the actual characteristics of multi-split conductors.

[0038] It is understandable that within the region of interest, the number of pixels in each effective wire cluster should be similar. For example, if the number of wire clusters after clustering is 4, the number of pixels in 3 of the wire clusters fluctuates around 200, while the number of pixels in another wire cluster is 500. In this case, it is obvious that the wire cluster with 500 pixels does not belong to the sub-wire and can be classified as a pseudo-cluster.

[0039] Optionally, if the number of wire clusters exceeds the first value, the number of pixels in each wire cluster is counted to determine the standard pixel number range. This range is the reasonable range of pixel numbers for a valid wire cluster. For each wire cluster, it is determined whether the number of pixels in the cluster belongs to the standard pixel number range. If the number of pixels is within the range, it means that the wire cluster conforms to the pixel distribution characteristics of the wire, and is determined to be a valid wire cluster and retained. If the number of pixels exceeds the standard pixel number range, it means that the cluster is a pseudo cluster and is directly deleted.

[0040] S140. Extract multi-dimensional feature parameters of effective conductor clusters, and generate visual annotations in aerial images based on the multi-dimensional feature parameters of effective conductor clusters.

[0041] Optionally, multi-dimensional feature parameters are a set of parameters used to quantify the features of effective conductor clusters. Visual annotations can intuitively display the position of effective conductor clusters and multi-dimensional feature parameters in aerial images. Visual annotations can include bounding boxes, text descriptions, pseudo-color annotations, etc., to facilitate inspection personnel to intuitively view the inspection results.

[0042] This involves extracting multi-dimensional feature parameters of effective traverse clusters and generating visual annotations in aerial images based on these parameters, including: Multi-dimensional feature parameters are extracted for each effective conductor cluster; among them, the multi-dimensional feature parameters include at least the number of pixels in the cluster, the minimum ordinate value, the maximum ordinate value, the center ordinate value, the average brightness, and the average depth. Based on the multi-dimensional feature parameters of the effective traverse clusters, bounding boxes, text descriptions of the feature parameters, and pseudo-color annotations are added to each effective traverse cluster in the aerial image.

[0043] Optionally, the number of pixels within a cluster is the total number of pixels contained within the effective traverse cluster; the minimum ordinate is the minimum ordinate of all pixels within the effective traverse cluster; the maximum ordinate is the maximum ordinate of all pixels within the effective traverse cluster; the center ordinate is the arithmetic mean of the minimum and maximum ordinate values ​​within the effective traverse cluster, reflecting the center position of the traverse cluster; the average brightness is the average brightness of all pixels within the effective traverse cluster, reflecting the brightness characteristics of the traverse; and the average depth is the average depth of all pixels within the effective traverse cluster, reflecting the relative distance between the UAV and the traverse.

[0044] Optionally, the extracted multi-dimensional feature parameters are combined with the position information of the effective conductor clusters. A bounding box is drawn for each effective conductor cluster in the original aerial image to accurately define the position of the conductor in the image. Different effective conductor clusters are marked with pseudo-color to distinguish different multi-splitting conductors by different colors, which facilitates quick identification and differentiation by inspection personnel. At the same time, text descriptions of feature parameters are added next to the bounding box to intuitively display core indicators such as the number of pixels in the cluster, center coordinates, and average brightness.

[0045] Furthermore, after extracting the multi-dimensional feature parameters of the effective conductor clusters and generating visual annotations in the aerial image based on these parameters, the process may further include: Based on the number of pixels within each effective conductor cluster in the aerial image, the vertical distance between the UAV and the multi-split conductor is determined when the UAV is acquiring the aerial image, and the vertical distance is sent to the UAV so that the UAV can adjust its motion parameters according to the vertical distance corresponding to each of the consecutive aerial images.

[0046] Optionally, the number of pixels in the aerial image of the multi-split duct is negatively correlated with the vertical distance between the UAV and the duct. That is, the closer the distance, the more pixels there are, and the farther the distance, the fewer pixels there are. By using a pre-calibrated mapping model of the relationship between the number of pixels and the vertical distance, the number of pixels within each effective duct cluster can be converted into the actual vertical distance between the UAV and the multi-split duct when the UAV is acquiring the current aerial image.

[0047] Furthermore, the calculated vertical distance is sent to the UAV's flight control system in real time. The flight control system can combine the vertical distances corresponding to multiple consecutive frames of aerial images to analyze the distance change trend between the UAV and the guide wire, and automatically adjust the UAV's motion parameters to maintain the optimal detection distance between the UAV and the guide wire. This ensures that the subsequently acquired aerial images are always at the best detection level, thereby improving the overall inspection efficiency.

[0048] The technical solution of this invention involves determining the region of interest (ROI) for a multi-splitting duct in an aerial image based on the motion state of the UAV, determining a brightness threshold corresponding to the aerial image based on the pixel brightness within the ROI, identifying multiple candidate duct pixels within the ROI based on the brightness threshold, and clustering these candidate pixels according to their positions to generate duct clusters. If the number of duct clusters generated after clustering exceeds a first value, valid duct clusters are identified within each cluster based on the number of pixels within each cluster. Multi-dimensional feature parameters of the valid duct clusters are then extracted, and based on... The multi-dimensional feature parameters of effective conductor clusters are used to generate visual annotations in aerial images. This method can dynamically determine the brightness threshold for the currently processed aerial image, solving the problem that fixed thresholds cannot adapt to different lighting conditions. It also solves the clustering and merging problem of multi-split conductors due to small spacing, ensuring accurate separation of multi-split conductors. For scenarios with too many conductor clusters after clustering, the number of pixels is used as the core screening criterion to quickly remove pseudo-clusters, avoiding interference from pseudo-clusters in the detection results and effectively improving the detection accuracy of conductor clusters. At the same time, it takes into account the real-time performance and robustness of the detection, and adapts to the computing power requirements of UAV embedded devices.

[0049] Example 2 Figure 2 This is a flowchart of a method for detecting multi-split conductors in transmission lines according to Embodiment 2 of the present invention. This embodiment, based on the above embodiments, specifically describes the method for detecting multi-split conductors in transmission lines. Figure 2 As shown, the method includes: S210. Based on the streaming video collected by the UAV, determine the flight direction of the UAV relative to the video image, and based on the flight direction and the typical distribution position of the conductor corresponding to the flight direction, determine the region of interest of the multi-split conductor in the aerial image.

[0050] Among them, aerial images are single-frame video images from streaming video.

[0051] S220. Based on the pixel brightness in the region of interest and the lower brightness limit constraint, determine the first lower brightness limit, and based on the first lower brightness limit, determine each effective pixel in the region of interest whose brightness is greater than the first lower brightness limit.

[0052] S230. Determine the brightness threshold corresponding to the aerial image based on the brightness of each effective pixel.

[0053] S240. Based on the brightness threshold, perform binarization segmentation on each pixel within the region of interest, and determine the pixels retained after binarization segmentation as candidate pixels for the conductor.

[0054] S250. Based on the vertical coordinate values ​​of each candidate pixel in the aerial image, sort the candidate pixels of the conductor, and according to the sorting results and the pre-set clustering interval, divide adjacent candidate pixels of the conductor with a vertical distance less than the clustering interval into the same conductor cluster.

[0055] S260. If the number of wire clusters generated after clustering exceeds the first value, then the valid wire clusters are identified in each wire cluster based on the number of pixels in each wire cluster.

[0056] If the number of wire clusters generated after clustering exceeds the first value, the standard pixel number range is determined based on the number of pixels in each wire cluster. Determine whether the number of pixels within the conductor cluster belongs to the standard pixel number range; If so, the wire cluster is determined as a valid wire cluster; If not, then determine that the wire cluster is a pseudo-cluster and delete the wire cluster.

[0057] Optionally, pseudo-clusters refer to false clusters of wires formed by image noise, background clutter, etc., without corresponding actual wires. Pseudo-clusters are the main factor leading to reduced detection accuracy.

[0058] Optionally, the standard pixel count range can be determined based on the number of pixels within each conductor cluster, which may include: Based on the maximum and minimum pixel counts of the conductor clusters and the preset target pixel count, multiple pixel count intervals are divided starting from the minimum pixel count until the maximum pixel count is covered. The pixel counts of each conductor cluster are then mapped to their respective pixel count intervals, and the target pixel count interval that corresponds to the most conductor clusters is determined as the standard pixel count interval.

[0059] In an optional example, if the target number of pixels is determined to be 30, the minimum pixel value is 190, and the maximum pixel value is 320, then the pixel number intervals can be divided as [190, 220], [220, 250], [250, 280], [280, 310], and [310, 340]. Then, based on the number of pixels in each wire cluster, the wire clusters are mapped to the pixel number intervals. Assuming that [250, 280] corresponds to the largest number of wire clusters, then [250, 280] is determined as the standard pixel number interval. If two adjacent pixel number intervals correspond to the same number of wire clusters, then the median value of the adjacent pixel number intervals can be taken to reconstruct the standard pixel number interval.

[0060] Optionally, the standard pixel count range can be determined based on the number of pixels within each conductor cluster, and may also include: Based on the number of pixels in each conductor cluster, the pixel mean and pixel standard deviation are calculated respectively. The difference between the pixel mean and the pixel standard deviation is taken as the minimum value of the standard pixel interval, and the sum of the pixel mean and the pixel standard deviation is taken as the maximum value of the standard pixel interval.

[0061] Furthermore, the method for detecting multi-split conductors in transmission lines may further include: If the number of wire clusters generated after clustering does not exceed the first value or no valid wire clusters are identified, the target adjustment ratio is determined according to the current identification round, and the second brightness lower limit is determined according to the target adjustment ratio and the first brightness lower limit. Based on the second brightness lower limit, valid pixels are re-determined within the region of interest, and valid wire clusters are identified in the next identification round based on the updated valid pixels, until the preset maximum identification round is reached or a valid wire cluster is identified.

[0062] Optionally, by designing the above-mentioned multi-level backoff verification mechanism, this application can solve the problem of missed detection of effective wire clusters in scenarios such as low light and complex backgrounds, and ensure the robustness of the detection algorithm.

[0063] Optionally, the target adjustment ratio is a brightness lower limit adjustment ratio set according to the identification round of the effective conductor cluster. It is used to relax the brightness screening conditions and generate a second brightness lower limit when no effective conductor cluster is identified, so as to realize multi-level backtracking verification.

[0064] Optionally, the second brightness lower limit refers to the brightness threshold value after the first brightness lower limit is adjusted by the target adjustment ratio. It is lower than the first brightness lower limit, which can include more low-brightness wire pixels and solve the problem of wires being missed in low-light scenes.

[0065] Optionally, if the number of wire clusters generated after clustering does not exceed the first value or no valid wire clusters are identified after step S260, it indicates that the current first brightness lower limit is set too strictly, resulting in a large number of low-brightness wire pixels being eliminated, triggering the valid wire cluster re-identification mechanism.

[0066] Furthermore, the maximum number of recognition rounds for the effective conductor cluster is preset, and the corresponding target adjustment ratio is determined based on the current recognition round; where the higher the round, the larger the adjustment ratio and the lower the brightness limit.

[0067] Furthermore, based on the first brightness lower limit, a second brightness lower limit is calculated and generated in combination with the target adjustment ratio. The second brightness lower limit is more lenient than the first brightness lower limit, and can include more low-brightness pixels.

[0068] Optionally, the second lower limit of brightness can be used as a new screening criterion to redefine the effective pixels in the region of interest, and steps S230-S260 can be repeated until the preset maximum number of recognition rounds is reached or an effective wire cluster is identified, so as to ensure that there are no missed detections of wires in low light scenes.

[0069] S270. Extract multi-dimensional feature parameters for each effective conductor cluster.

[0070] Among them, the multi-dimensional feature parameters include at least the number of pixels within a cluster, the minimum ordinate value, the maximum ordinate value, the center ordinate value, the average brightness, and the average depth.

[0071] S280. Based on the multi-dimensional feature parameters of the effective traverse clusters, add bounding boxes, text descriptions of the feature parameters, and pseudo-color annotations to each effective traverse cluster in the aerial image.

[0072] The technical solution of this invention involves determining the region of interest (ROI) for a multi-splitting duct in an aerial image based on the motion state of the UAV, determining a brightness threshold corresponding to the aerial image based on the pixel brightness within the ROI, identifying multiple candidate duct pixels within the ROI based on the brightness threshold, and clustering these candidate pixels according to their positions to generate duct clusters. If the number of duct clusters generated after clustering exceeds a first value, valid duct clusters are identified within each cluster based on the number of pixels within each cluster. Multi-dimensional feature parameters of the valid duct clusters are then extracted, and based on... The multi-dimensional feature parameters of effective conductor clusters are used to generate visual annotations in aerial images. This method can dynamically determine the brightness threshold for the currently processed aerial image, solving the problem that fixed thresholds cannot adapt to different lighting conditions. It also solves the clustering and merging problem of multi-split conductors due to small spacing, ensuring accurate separation of multi-split conductors. For scenarios with too many conductor clusters after clustering, the number of pixels is used as the core screening criterion to quickly remove pseudo-clusters, avoiding interference from pseudo-clusters in the detection results and effectively improving the detection accuracy of conductor clusters. At the same time, it takes into account the real-time performance and robustness of the detection, and adapts to the computing power requirements of UAV embedded devices.

[0073] Example 3 Figure 3 This is a schematic diagram of the structure of a multi-split conductor detection device for power transmission lines provided in Embodiment 3 of the present invention. Figure 3 As shown, the device includes: a brightness threshold determination module 310, a conductor cluster generation module 320, an effective conductor cluster identification module 330, and a visualization processing module 340.

[0074] The brightness threshold determination module 310 is used to determine the region of interest of the multi-split conductor in the aerial image based on the motion state of the UAV, and to determine the brightness threshold corresponding to the aerial image based on the pixel brightness in the region of interest.

[0075] The conductor cluster generation module 320 is used to determine multiple conductor candidate pixels in the region of interest according to the brightness threshold, and to perform clustering processing on the conductor candidate pixels according to the position of each conductor candidate pixel to generate a conductor cluster.

[0076] The effective conductor cluster identification module 330 is used to identify effective conductor clusters in each conductor cluster based on the number of pixels in each conductor cluster if the number of conductor clusters generated after clustering exceeds a first value.

[0077] The visualization processing module 340 is used to extract multi-dimensional feature parameters of effective conductor clusters and generate visual annotations in aerial images based on the multi-dimensional feature parameters of effective conductor clusters.

[0078] The technical solution of this invention involves determining the region of interest (ROI) for a multi-splitting duct in an aerial image based on the motion state of the UAV, determining a brightness threshold corresponding to the aerial image based on the pixel brightness within the ROI, identifying multiple candidate duct pixels within the ROI based on the brightness threshold, and clustering these candidate pixels according to their positions to generate duct clusters. If the number of duct clusters generated after clustering exceeds a first value, valid duct clusters are identified within each cluster based on the number of pixels within each cluster. Multi-dimensional feature parameters of the valid duct clusters are then extracted, and based on... The multi-dimensional feature parameters of effective conductor clusters are used to generate visual annotations in aerial images. This method can dynamically determine the brightness threshold for the currently processed aerial image, solving the problem that fixed thresholds cannot adapt to different lighting conditions. It also solves the clustering and merging problem of multi-split conductors due to small spacing, ensuring accurate separation of multi-split conductors. For scenarios with too many conductor clusters after clustering, the number of pixels is used as the core screening criterion to quickly remove pseudo-clusters, avoiding interference from pseudo-clusters in the detection results and effectively improving the detection accuracy of conductor clusters. At the same time, it takes into account the real-time performance and robustness of the detection, and adapts to the computing power requirements of UAV embedded devices.

[0079] Based on the above embodiments, the brightness threshold determination module 310 can be specifically used for: Based on the streaming video captured by the UAV, the flight direction of the UAV relative to the video frame is determined, and based on the flight direction and the typical distribution position of the conductors corresponding to the flight direction, the region of interest of the multi-split conductors is determined in the aerial image; wherein, the aerial image is a single frame video image in the streaming video. Based on the pixel brightness within the region of interest and the lower limit constraint for brightness, a first lower limit for brightness is determined, and based on the first lower limit for brightness, each effective pixel within the region of interest whose brightness is greater than the first lower limit for brightness is determined. Based on the brightness of each effective pixel, a brightness threshold corresponding to the aerial image is determined.

[0080] Based on the above embodiments, the wire cluster generation module 320 can be specifically used for: Based on the brightness threshold, each pixel in the region of interest is binarized and segmented, and the pixels retained after binarization are determined as candidate pixels for the conductor. Based on the vertical coordinate values ​​of each candidate pixel in the aerial image, the candidate pixels are sorted, and based on the sorting results and the pre-set clustering interval, adjacent candidate pixels with a vertical distance less than the clustering interval are divided into the same candidate pixel cluster.

[0081] Based on the above embodiments, the effective conductor cluster identification module 330 can be specifically used for: If the number of wire clusters generated after clustering exceeds the first value, then the standard pixel number range is determined based on the number of pixels in each wire cluster. Determine whether the number of pixels within the conductor cluster belongs to the standard pixel number range; If so, the wire cluster is determined as a valid wire cluster; If not, then determine that the wire cluster is a pseudo-cluster and delete the wire cluster.

[0082] Based on the above embodiments, a rollback verification module may also be included, used for: If the number of wire clusters generated after clustering does not exceed the first value or no valid wire clusters are identified, the target adjustment ratio is determined according to the current identification round, and the second brightness lower limit is determined according to the target adjustment ratio and the first brightness lower limit. Based on the second brightness lower limit, valid pixels are re-determined within the region of interest, and valid wire clusters are identified in the next identification round based on the updated valid pixels, until the preset maximum identification round is reached or a valid wire cluster is identified.

[0083] Based on the above embodiments, the visualization processing module 340 can be specifically used for: Multi-dimensional feature parameters are extracted for each effective conductor cluster; among them, the multi-dimensional feature parameters include at least the number of pixels in the cluster, the minimum ordinate value, the maximum ordinate value, the center ordinate value, the average brightness, and the average depth. Based on the multi-dimensional feature parameters of the effective traverse clusters, bounding boxes, text descriptions of the feature parameters, and pseudo-color annotations are added to each effective traverse cluster in the aerial image.

[0084] Based on the above embodiments, a vertical distance detection module may also be included, for: Based on the number of pixels within each effective conductor cluster in the aerial image, the vertical distance between the UAV and the multi-split conductor is determined when the UAV is acquiring the aerial image, and the vertical distance is sent to the UAV so that the UAV can adjust its motion parameters according to the vertical distance corresponding to each of the consecutive aerial images.

[0085] The transmission line multi-split conductor detection device provided in this embodiment of the invention can execute the transmission line multi-split conductor detection method provided in any embodiment of the invention, and has the corresponding functional modules and beneficial effects of the method.

[0086] Example 4 Figure 4 A schematic diagram of an electronic device 10, which can be used to implement embodiments of the present invention, is shown. The electronic device is intended to represent various forms of digital computers, such as laptop computers, desktop computers, workstations, personal digital assistants, servers, blade servers, mainframe computers, and other suitable computers. The electronic device can also represent various forms of mobile devices, such as personal digital processors, cellular phones, smartphones, wearable devices (e.g., helmets, glasses, watches, etc.), and other similar computing devices. The components shown herein, their connections and relationships, and their functions are merely illustrative and are not intended to limit the implementation of the invention described and / or claimed herein.

[0087] like Figure 4 As shown, the electronic device 10 includes at least one processor 11 and a memory, such as a read-only memory (ROM) 12 or a random access memory (RAM) 13, communicatively connected to the at least one processor 11. The memory stores computer programs executable by the at least one processor. The processor 11 can perform various appropriate actions and processes based on the computer program stored in the ROM 12 or loaded from storage unit 18 into the RAM 13. The RAM 13 can also store various programs and data required for the operation of the electronic device 10. The processor 11, ROM 12, and RAM 13 are interconnected via a bus 14. An input / output (I / O) interface 15 is also connected to the bus 14.

[0088] Multiple components in electronic device 10 are connected to I / O interface 15, including: input unit 16, such as keyboard, mouse, etc.; output unit 17, such as various types of displays, speakers, etc.; storage unit 18, such as disk, optical disk, etc.; and communication unit 19, such as network card, modem, wireless transceiver, etc. Communication unit 19 allows electronic device 10 to exchange information / data with other devices through computer networks such as the Internet and / or various telecommunications networks.

[0089] Processor 11 can be various general-purpose and / or special-purpose processing components with processing and computing capabilities. Some examples of processor 11 include, but are not limited to, central processing unit (CPU), graphics processing unit (GPU), various special-purpose artificial intelligence (AI) computing chips, various processors running machine learning model algorithms, digital signal processors (DSPs), and any suitable processor, controller, microcontroller, etc. Processor 11 performs the various methods and processes described above, such as the method for detecting multi-split conductors in transmission lines according to embodiments of the present invention. That is: Based on the movement state of the UAV, the region of interest for the multi-split conductor is determined in the aerial image, and the brightness threshold corresponding to the aerial image is determined based on the pixel brightness within the region of interest. Based on the brightness threshold, multiple candidate pixels for conductors are determined within the region of interest, and the candidate pixels for conductors are clustered according to the position of each candidate pixel to generate a conductor cluster. If the number of wire clusters generated after clustering exceeds the first value, then the valid wire clusters are identified in each wire cluster based on the number of pixels in each wire cluster. Extract multi-dimensional feature parameters of effective conductor clusters, and generate visual annotations in aerial images based on these parameters.

[0090] In some embodiments, the transmission line multi-branch conductor detection method may be implemented as a computer program tangibly contained in a computer-readable storage medium, such as storage unit 18. In some embodiments, part or all of the computer program may be loaded and / or installed on electronic device 10 via ROM 12 and / or communication unit 19. When the computer program is loaded into RAM 13 and executed by processor 11, one or more steps of the transmission line multi-branch conductor detection method described above may be performed. Alternatively, in other embodiments, processor 11 may be configured to perform the transmission line multi-branch conductor detection method by any other suitable means (e.g., by means of firmware).

[0091] Various embodiments of the systems and techniques described above herein can be implemented in digital electronic circuit systems, integrated circuit systems, field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), computer hardware, firmware, software, and / or combinations thereof. These various embodiments may include implementations in one or more computer programs that can be executed and / or interpreted on a programmable system including at least one programmable processor, which may be a dedicated or general-purpose programmable processor, capable of receiving data and instructions from a storage system, at least one input device, and at least one output device, and transmitting data and instructions to the storage system, the at least one input device, and the at least one output device.

[0092] Computer programs used to implement the methods of the present invention may be written in any combination of one or more programming languages. These computer programs may be provided to a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing device, such that when executed by the processor, the computer programs cause the functions / operations specified in the flowcharts and / or block diagrams to be performed. The computer programs may be executed entirely on a machine, partially on a machine, or as a standalone software package, partially on a machine and partially on a remote machine, or entirely on a remote machine or server.

[0093] In the context of this invention, a computer-readable storage medium can be a tangible medium that may contain or store a computer program for use by or in conjunction with an instruction execution system, apparatus, or device. A computer-readable storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination thereof. Alternatively, a computer-readable storage medium may be a machine-readable signal medium. More specific examples of machine-readable storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination thereof.

[0094] To provide interaction with a user, the systems and techniques described herein can be implemented on an electronic device having: a display device (e.g., a CRT (cathode ray tube) or LCD (liquid crystal display) monitor) for displaying information to the user; and a keyboard and pointing device (e.g., a mouse or trackball) through which the user provides input to the electronic device. Other types of devices can also be used to provide interaction with the user; for example, feedback provided to the user can be any form of sensory feedback (e.g., visual feedback, auditory feedback, or tactile feedback); and input from the user can be received in any form (including sound input, voice input, or tactile input).

[0095] The systems and technologies described herein can be implemented in computing systems that include backend components (e.g., as data servers), or middleware components (e.g., application servers), or frontend components (e.g., user computers with graphical user interfaces or web browsers through which users can interact with implementations of the systems and technologies described herein), or any combination of such backend, middleware, or frontend components. The components of the system can be interconnected via digital data communication of any form or medium (e.g., communication networks). Examples of communication networks include local area networks (LANs), wide area networks (WANs), blockchain networks, and the Internet.

[0096] A computing system can include clients and servers. Clients and servers are generally located far apart and typically interact through communication networks. The client-server relationship is created by computer programs running on the respective computers and having a client-server relationship with each other. The server can be a cloud server, also known as a cloud computing server or cloud host, which is a hosting product within the cloud computing service system to address the shortcomings of traditional physical hosts and VPS services, such as high management difficulty and weak business scalability.

[0097] It should be understood that the various forms of processes shown above can be used, with steps reordered, added, or deleted. For example, the steps described in this invention can be executed in parallel, sequentially, or in different orders, as long as the desired result of the technical solution of this invention can be achieved, and this is not limited herein.

[0098] The specific embodiments described above do not constitute a limitation on the scope of protection of this invention. Those skilled in the art should understand that various modifications, combinations, sub-combinations, and substitutions can be made according to design requirements and other factors. Any modifications, equivalent substitutions, and improvements made within the spirit and principles of this invention should be included within the scope of protection of this invention.

Claims

1. A method for detecting multi-split conductors in transmission lines, characterized in that, include: Based on the movement state of the UAV, the region of interest for the multi-split conductor is determined in the aerial image, and the brightness threshold corresponding to the aerial image is determined based on the pixel brightness within the region of interest. Based on the brightness threshold, multiple candidate pixels for conductors are determined within the region of interest, and the candidate pixels for conductors are clustered according to the position of each candidate pixel to generate a conductor cluster. If the number of wire clusters generated after clustering exceeds the first value, then the valid wire clusters are identified in each wire cluster based on the number of pixels in each wire cluster. Extract multi-dimensional feature parameters of effective conductor clusters, and generate visual annotations in aerial images based on the multi-dimensional feature parameters of effective conductor clusters; If the number of wire clusters generated after clustering exceeds the first value, then valid wire clusters are identified in each wire cluster based on the number of pixels within each wire cluster, including: If the number of wire clusters generated after clustering exceeds the first value, then the standard pixel number range is determined based on the number of pixels in each wire cluster; wherein, the first value is set according to the number of multiple sub-wires of the same type contained in the multi-split wire. Determine whether the number of pixels within the conductor cluster belongs to the standard pixel number range; If so, the wire cluster is determined as a valid wire cluster; If not, then determine that the wire cluster is a pseudo cluster and delete the wire cluster; Specifically, the standard pixel count range is determined based on the number of pixels within each conductor cluster, including: Based on the maximum and minimum pixel counts of the conductor clusters and the preset target pixel count, multiple pixel count intervals are divided starting from the minimum pixel count, until the maximum pixel count is covered. The pixel count of each conductor cluster is then mapped to each pixel count interval, and the target pixel count interval that corresponds to the most conductor clusters is determined as the standard pixel count interval. If two adjacent pixel count intervals correspond to the same number of conductor clusters, the median of the adjacent pixel count intervals is taken to reconstruct the standard pixel count interval.

2. The method according to claim 1, characterized in that, Based on the drone's motion state, the region of interest (ROI) for the multi-split duct is determined in the aerial image, and a brightness threshold corresponding to the aerial image is determined based on the pixel brightness within the ROI, including: Based on the streaming video captured by the UAV, the flight direction of the UAV relative to the video frame is determined, and based on the flight direction and the typical distribution position of the conductors corresponding to the flight direction, the region of interest of the multi-split conductors is determined in the aerial image; wherein, the aerial image is a single frame video image in the streaming video. Based on the pixel brightness within the region of interest and the lower limit constraint for brightness, a first lower limit for brightness is determined, and based on the first lower limit for brightness, each effective pixel within the region of interest whose brightness is greater than the first lower limit for brightness is determined. Based on the brightness of each effective pixel, a brightness threshold corresponding to the aerial image is determined.

3. The method according to claim 1, characterized in that, Based on the brightness threshold, multiple candidate pixels for conductors are determined within the region of interest, and based on the position of each candidate pixel, the candidate pixels are clustered to generate conductor clusters, including: Based on the brightness threshold, each pixel in the region of interest is binarized and segmented, and the pixels retained after binarization are determined as candidate pixels for the conductor. Based on the vertical coordinate values ​​of each candidate pixel in the aerial image, the candidate pixels are sorted, and based on the sorting results and the pre-set clustering interval, adjacent candidate pixels with a vertical distance less than the clustering interval are divided into the same candidate pixel cluster.

4. The method according to claim 2, characterized in that, Also includes: If the number of wire clusters generated after clustering does not exceed the first value or no valid wire clusters are identified, the target adjustment ratio is determined according to the current identification round, and the second brightness lower limit is determined according to the target adjustment ratio and the first brightness lower limit. Based on the second lower limit of brightness, valid pixels are re-determined within the region of interest, and valid wire clusters are identified in the next identification round based on the updated valid pixels, until the preset maximum identification round is reached or a valid wire cluster is identified.

5. The method according to claim 1, characterized in that, Extract multi-dimensional feature parameters of effective conductor clusters, and generate visual annotations in aerial images based on these parameters, including: Multi-dimensional feature parameters are extracted for each effective conductor cluster; among them, the multi-dimensional feature parameters include at least the number of pixels in the cluster, the minimum ordinate value, the maximum ordinate value, the center ordinate value, the average brightness, and the average depth. Based on the multi-dimensional feature parameters of the effective traverse clusters, bounding boxes, text descriptions of the feature parameters, and pseudo-color annotations are added to each effective traverse cluster in the aerial image.

6. The method according to claim 5, characterized in that, After extracting the multi-dimensional feature parameters of the effective traverse clusters and generating visual annotations in the aerial images based on these parameters, the process also includes: Based on the number of pixels within each effective conductor cluster in the aerial image, the vertical distance between the UAV and the multi-split conductor is determined when the UAV is acquiring aerial images. The vertical distance is then sent to the UAV so that the UAV can adjust its motion parameters according to the vertical distance corresponding to each of the consecutive aerial images.

7. A device for detecting multi-split conductors in transmission lines, characterized in that, include: The brightness threshold determination module is used to determine the region of interest of the multi-split conductor in the aerial image based on the motion state of the UAV, and to determine the brightness threshold corresponding to the aerial image based on the pixel brightness in the region of interest. The conductor cluster generation module is used to determine multiple conductor candidate pixels in the region of interest according to the brightness threshold, and to perform clustering processing on the conductor candidate pixels according to the position of each conductor candidate pixel to generate a conductor cluster. The effective conductor cluster identification module is used to identify effective conductor clusters in each conductor cluster based on the number of pixels in each conductor cluster if the number of conductor clusters generated after clustering exceeds a first value. The visualization processing module is used to extract multi-dimensional feature parameters of effective conductor clusters and generate visual annotations in aerial images based on the multi-dimensional feature parameters of effective conductor clusters. The effective conductor cluster identification module is specifically used for: If the number of wire clusters generated after clustering exceeds the first value, then the standard pixel number range is determined based on the number of pixels in each wire cluster; wherein, the first value is set according to the number of multiple sub-wires of the same type contained in the multi-split wire. Determine whether the number of pixels within the conductor cluster belongs to the standard pixel number range; If so, the wire cluster is determined as a valid wire cluster; If not, then determine that the wire cluster is a pseudo cluster and delete the wire cluster; Specifically, the standard pixel count range is determined based on the number of pixels within each conductor cluster, including: Based on the maximum and minimum pixel counts of the conductor clusters and the preset target pixel count, multiple pixel count intervals are divided starting from the minimum pixel count, until the maximum pixel count is covered. The pixel count of each conductor cluster is then mapped to each pixel count interval, and the target pixel count interval that corresponds to the most conductor clusters is determined as the standard pixel count interval. If two adjacent pixel count intervals correspond to the same number of conductor clusters, the median of the adjacent pixel count intervals is taken to reconstruct the standard pixel count interval.

8. An electronic device, characterized in that, The electronic device includes: At least one processor; and A memory communicatively connected to the at least one processor; wherein, The memory stores a computer program that can be executed by the at least one processor, the computer program being executed by the at least one processor to enable the at least one processor to perform the transmission line multi-split conductor detection method according to any one of claims 1-6.

9. A computer-readable storage medium, characterized in that, The computer-readable storage medium stores computer instructions that cause a processor to execute the method for detecting multi-split conductors in transmission lines as described in any one of claims 1-6.

Images