Power transmission line point cloud parameter extraction method and system

Abstract

The invention relates to the technical field of point cloud data processing, and discloses a power transmission line point cloud parameter extraction method and system, and the method comprises the steps: carrying out the point cloud semantic segmentation of point cloud data, so as to determine the point cloud data corresponding to a plurality of different equipment types, and carrying out the clustering processing of the point cloud data of a power transmission tower, so as to obtain the point cloud parameters of a power transmission line. Performing point cloud division according to the point cloud data set of the power transmission tower obtained by clustering, dividing point cloud data of a power transmission line associated with the power transmission tower, and updating the point cloud data to the point cloud data set of the power transmission tower, thereby accurately covering the point cloud data of the power transmission line, and extracting key point cloud parameters of the power transmission tower. Therefore, the problems of information loss, redundant and complex calculation content and overlarge memory occupation in the calculation process are avoided, and the point cloud parameter extraction precision and efficiency are improved.

Description

Technical Field

[0001] The present invention relates to the technical field of point cloud data processing, and in particular to a method and system for extracting point cloud parameters of a power transmission line. Background Art

[0002] The construction of digital power grids has put forward higher requirements for the automation and intelligence of lightning protection for transmission lines. In order to realize automated and intelligent lightning protection, it is necessary to first digitally model the transmission lines and towers. This process usually consumes a lot of manpower and material resources, significantly reducing the efficiency of lightning protection analysis. On the one hand, the existing lightning protection analysis programs are mostly stand-alone software, which usually requires manual input of a large number of line parameters to carry out lightning protection performance evaluation, which brings huge workload pressure to the staff. In recent years, with the development of laser radar technology, laser radar technology has been widely used in unmanned driving, transmission line inspection and other fields. Through laser radar technology, the three-dimensional coordinate information of the surface of an object can be obtained, that is, laser point cloud data. Therefore, airborne laser radar can be used to obtain point cloud data of the transmission line, and then the line and tower structure parameters can be automatically extracted from the laser point cloud data, thereby realizing digital modeling of the transmission line.

[0003] The existing point cloud parameter extraction method for power transmission lines mainly converts the original three-dimensional point cloud data into two-dimensional images for processing. However, this point cloud parameter extraction method is prone to problems such as information loss, redundant and complex calculation content, and excessive memory usage during the calculation process, resulting in poor point cloud parameter extraction accuracy and efficiency. Summary of the invention

[0004] The present invention provides a method and system for extracting point cloud parameters of a power transmission line, which solves the technical problems that the existing point cloud parameter extraction methods for power transmission lines are prone to information loss, redundant and complex calculation content, and excessive memory usage during the calculation process, resulting in poor point cloud parameter extraction accuracy and efficiency.

[0005] In view of this, a first aspect of the present invention provides a method for extracting point cloud parameters of a transmission line, comprising:

[0006] Preprocessing of point cloud data of transmission lines;

[0007] Performing point cloud semantic segmentation on the preprocessed point cloud data, wherein the point cloud semantic segmentation is used to determine point cloud data corresponding to a plurality of different device types;

[0008] Determine the point cloud data of the transmission tower according to the point cloud semantic segmentation result, perform clustering processing on the point cloud data of the transmission tower, and obtain a plurality of different point cloud data sets of the transmission tower;

[0009] Performing point cloud division according to the point cloud data set of the transmission tower, dividing the point cloud data of the transmission line associated with the transmission tower, and updating the point cloud data of the transmission line to the point cloud data set of the transmission tower;

[0010] The key point cloud parameters of the transmission tower are extracted according to the updated point cloud data set of the transmission tower.

[0011] Preferably, the preprocessing includes point cloud segmentation processing and point cloud denoising processing.

[0012] Preferably, the step of segmenting the point cloud comprises:

[0013] Determining boundary point cloud data according to the point cloud data of the transmission line;

[0014] Determine, according to the boundary point cloud data, a first minimum enclosing rectangle of the point cloud data of the transmission line as a boundary box of the transmission line;

[0015] Point cloud segmentation is performed according to the maximum side length of the boundary box of the power transmission line according to a preset segment moving step length, and the point cloud data of the power transmission line is segmented into point cloud data of a plurality of segments.

[0016] Preferably, the step of performing point cloud division according to the point cloud data set of the transmission tower, dividing the point cloud data of the transmission line associated with the transmission tower, and updating the point cloud data of the transmission line to the point cloud data set of the transmission tower specifically includes:

[0017] Performing a mean operation on the point cloud data set of the transmission tower to obtain the centroid point cloud coordinates of the transmission tower, and determining the point cloud centroid distance between each point cloud data in the point cloud data set of the transmission tower and the centroid point cloud coordinates;

[0018] Filter out a plurality of point cloud data whose point cloud gravity center distance is less than a preset distance threshold, and construct a tower gravity center point cloud data set;

[0019] Determine a second minimum enclosing rectangle based on the tower gravity center point cloud data set;

[0020] Determine the width of the tower according to the shortest side of the second minimum enclosing rectangle;

[0021] Taking the point cloud coordinates of the center of gravity of the transmission tower as the origin and the width of the tower at a preset multiple as the radius, a circular area is determined, and all point cloud data within the circular area are determined to be point cloud data of the transmission line associated with the transmission tower;

[0022] The point cloud data of the transmission line associated with the transmission tower is updated into the point cloud data set of the transmission tower.

[0023] Preferably, the step of extracting key point cloud parameters of the transmission tower according to the updated point cloud data set of the transmission tower specifically includes:

[0024] Projecting the updated point cloud data set of the transmission tower onto a two-dimensional coordinate plane to obtain a two-dimensional coordinate point set;

[0025] Processing the two-dimensional coordinate point set based on the PCA algorithm to obtain the main plane vector direction;

[0026] Rotating and projecting the point cloud data set of the transmission tower into a grid in the direction of the main plane vector to generate a two-dimensional image;

[0027] Performing edge smoothing processing on the two-dimensional image;

[0028] The Canny edge detection operator is used to perform edge detection on the two-dimensional image after edge smoothing, and the maximum outer contour is extracted according to the edge detection result;

[0029] Using an LSD straight line detection algorithm to perform straight line segment detection on the maximum outer contour, and determining a plurality of straight line segments whose lengths are less than a preset length threshold according to the straight line segment detection results;

[0030] The screened multiple straight line segments are merged, and the intersection points between the merged multiple straight line segments are extracted as the key corner points of the transmission tower.

[0031] Preferably, the main plane vector direction includes a transverse vector direction of the tower main plane projection and a vertical vector direction of the tower main plane projection;

[0032] After the step of processing the two-dimensional coordinate point set based on the PCA algorithm to obtain the main plane vector direction, the method further includes:

[0033] Determine whether the angle between the vector direction between the center point of the transmission tower and the center point of its adjacent transmission tower and the horizontal vector direction of the main plane projection of the tower is an acute angle;

[0034] If it is determined that the angle is not an acute angle, the transverse vector direction of the projection of the main plane of the tower is rotated to 180° along the Z-axis direction with the center of gravity of the tower as the origin, and the transverse vector direction of the projection of the main plane of the tower obtained after the rotation is updated to the new transverse vector direction of the projection of the main plane of the tower.

[0035] Preferably, after the step of merging the screened multiple straight line segments and extracting the intersection points between the merged multiple straight line segments as the key corner points of the transmission tower, the method further comprises:

[0036] Determine the XZ two-dimensional coordinates of the key corner points of the transmission tower according to the intersection points;

[0037] Projecting the point cloud data set of the transmission tower onto a two-dimensional plane in the direction of the side view vector, and determining the maximum coordinate value in the Y-axis direction corresponding to the Z-axis coordinate value of the key corner point of the transmission tower based on the projection of the transmission tower onto the two-dimensional plane in the direction of the side view vector;

[0038] The three-dimensional coordinates of the key corner points of the transmission tower are determined according to the maximum coordinate value in the Y-axis direction and the XZ two-dimensional coordinates of the key corner points of the transmission tower.

[0039] Preferably, after the step of extracting key point cloud parameters of the transmission tower according to the updated point cloud data set of the transmission tower, the method further comprises:

[0040] According to the point cloud semantic segmentation result, the point cloud data of the conductor is rotated to the main plane vector direction, an O-XYZ three-dimensional coordinate system is established with the center of gravity of the transmission tower as the origin, and the point cloud data of the conductor is divided into point cloud data of the conductors on both sides through the XOZ plane;

[0041] The point cloud data of the conductors on both sides are clustered to obtain multiple single-phase conductors;

[0042] Determine a coordinate sequence data set of each single-phase conductor according to the coordinate position distribution of each single-phase conductor in the O-XYZ three-dimensional coordinate system;

[0043] Using the RANSAC plane fitting algorithm to fit the coordinate sequence data set of the single-phase conductor to obtain the plane equation of the single-phase conductor;

[0044] The RANSAC line fitting algorithm is used to fit the coordinate sequence data set of the single-phase conductor to obtain the catenary equation of the single-phase conductor.

[0045] Preferably, the method further comprises:

[0046] Extract GIS data, front span and line turning angle based on the updated point cloud data set of transmission towers.

[0047] In a second aspect, the present invention further provides a transmission line point cloud parameter extraction system, comprising:

[0048] A preprocessing module, used for preprocessing point cloud data of transmission lines;

[0049] A semantic segmentation module, used to perform point cloud semantic segmentation on the preprocessed point cloud data, wherein the point cloud semantic segmentation is used to determine point cloud data corresponding to a plurality of different device types;

[0050] A point cloud clustering module is used to determine the point cloud data of the transmission tower according to the point cloud semantic segmentation result, and perform clustering processing on the point cloud data of the transmission tower to obtain a plurality of different point cloud data sets of the transmission tower;

[0051] A point cloud division module, used to perform point cloud division according to the point cloud data set of the transmission tower, divide the point cloud data of the transmission line associated with the transmission tower, and update the point cloud data of the transmission line to the point cloud data set of the transmission tower;

[0052] The point cloud parameter extraction module is used to extract key point cloud parameters of the transmission tower according to the updated point cloud data set of the transmission tower.

[0053] It can be seen from the above technical solutions that the present invention has the following advantages:

[0054] The present invention performs point cloud semantic segmentation on point cloud data to determine point cloud data corresponding to multiple different device types, clusters the point cloud data of the transmission tower, divides the point cloud according to the point cloud data set of the transmission tower obtained by clustering, divides the point cloud data of the transmission line associated with the transmission tower, and updates the point cloud data set of the transmission tower, thereby accurately covering the point cloud data of the transmission line and extracting key point cloud parameters of the transmission tower, thereby avoiding the problems of lost information, redundant and complex calculation content, and excessive memory usage in the calculation process, and improving the accuracy and efficiency of point cloud parameter extraction. BRIEF DESCRIPTION OF THE DRAWINGS

[0055] Figure 1 A flow chart of a method for extracting point cloud parameters of a power transmission line provided by an embodiment of the present invention;

[0056] Figure 2 A flowchart of point cloud segmentation processing provided by an embodiment of the present invention;

[0057] Figure 3 A flowchart of step S4 provided in an embodiment of the present invention;

[0058] Figure 4 A flowchart of step S5 provided in an embodiment of the present invention;

[0059] Figure 5 A side view of a tower provided by an embodiment of the present invention;

[0060] Figure 6 A flow chart of the catenary equation fitting of a conductor provided in an embodiment of the present invention;

[0061] Figure 7 A schematic diagram of the structure of a transmission line point cloud parameter extraction system provided in an embodiment of the present invention. DETAILED DESCRIPTION

[0062] In order to enable those skilled in the art to better understand the scheme of the present invention, the technical scheme in the embodiments of the present invention will be clearly and completely described below in conjunction with the drawings in the embodiments of the present invention. Obviously, the 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.

[0063] It should be noted that the terms "first", "second", etc. in the specification and claims of the present invention and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchanged where appropriate, so that the embodiments of the present invention described herein can be implemented in an order other than those illustrated or described herein. In addition, the terms "including" and "having" and any variations thereof are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device that includes a series of steps or units is not necessarily limited to those steps or units that are clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0064] The existing point cloud parameter extraction method for power transmission lines mainly converts the original three-dimensional point cloud data into two-dimensional images for processing. However, this point cloud parameter extraction method is prone to problems such as information loss, redundant and complex calculation content, and excessive memory usage during the calculation process, resulting in poor point cloud parameter extraction accuracy and efficiency.

[0065] In view of this, an embodiment of the present invention provides a method for extracting point cloud parameters of a transmission line. The embodiment of the present invention can be applicable to the case of extracting point cloud parameters of a transmission line. The method can be executed by a transmission line point cloud parameter extraction device. The transmission line point cloud parameter extraction device can be implemented in the form of hardware and / or software. The transmission line point cloud parameter extraction device can be configured in a computer device.

[0066] like Figure 1 As shown, Figure 1 The flowchart of a method for extracting point cloud parameters of a power transmission line provided by the present invention is illustrated.

[0067] The present invention provides a method for extracting point cloud parameters of a power transmission line, comprising steps S1 to S5:

[0068] Step S1: preprocessing the point cloud data of the transmission line.

[0069] Among them, the point cloud data of the transmission line is obtained by scanning along the transmission line with a drone carrying an airborne lidar, which includes point sets such as transmission lines, towers, insulators, and nearby ground, trees, and buildings.

[0070] Among them, the preprocessing includes point cloud segmentation processing and point cloud denoising processing.

[0071] Since the length of the line scanned by the laser radar during a single power line inspection is often tens of kilometers or more, the point cloud data file usually reaches several G bytes to several T bytes. In order to reduce the memory pressure and computing pressure, this embodiment performs point cloud segmentation processing on the point cloud data of the power transmission line. Specifically, Figure 2 As shown, the steps of point cloud segmentation processing include:

[0072] Step S101: Determine boundary point cloud data according to point cloud data of a power transmission line.

[0073] The boundary point cloud data includes the maximum and minimum values ​​of the point cloud data of the transmission line in the X-axis direction, and the maximum and minimum values ​​in the Y-axis direction.

[0074] Step S102: determining a first minimum enclosing rectangle of the point cloud data of the power transmission line according to the boundary point cloud data as a boundary box of the power transmission line.

[0075] Step S103: segmenting the point cloud according to the maximum side length of the boundary box of the power transmission line according to the preset segment moving step length, and segmenting the point cloud data of the power transmission line into point cloud data of a plurality of segments.

[0076] Among them, since the section may divide the transmission tower into two parts during the segmentation process, an operation similar to image convolution is used to segment the point cloud data, and the segment movement step size can be set to L-ΔL, where L is the average distance threshold of the tower and ΔL is 3 times the maximum tower width.

[0077] The average distance threshold of the towers can be obtained by dividing the total length (the maximum side length of the bounding box) by the number of towers.

[0078] For point cloud denoising, due to the influence of noise points (flying birds, dust, tall trees, etc.), hardware imaging quality, jitter, light and other factors in the laser radar scanning scene, the original tower point cloud often has a lot of noise, which greatly affects the point cloud quality and increases the difficulty of subsequent point cloud processing. Therefore, it is necessary to denoise the point cloud. The present invention performs denoising based on the point density method, by calculating the k nearest neighbors of each point, and then calculating the average distance of the k nearest neighbors of the point And filter out point clouds whose average distance is higher than the preset distance threshold.

[0079] Step S2: performing point cloud semantic segmentation on the preprocessed point cloud data. The point cloud semantic segmentation is used to determine the point cloud data corresponding to multiple different device types.

[0080] Among them, a semantic segmentation model based on deep learning can be used to perform semantic segmentation on point cloud data. Point cloud semantic segmentation can determine the point cloud data corresponding to multiple different equipment types. The equipment types include but are not limited to insulator string points, ground points, transmission line points and transmission tower points.

[0081] Since there are some misclassifications in the semantic segmentation results, the point cloud data needs to be denoised and filtered again. For the semantically segmented point cloud, the point cloud data is clustered according to the semantic segmentation labels to remove the misclassified data points.

[0082] Exemplarily, the DBSCAN clustering method can be used to perform DBSCAN clustering on each type of point set, and points with a number of points below a threshold are regarded as noise, thereby filtering out noise points and removing misclassified data points.

[0083] Step S3: determine the point cloud data of the transmission tower according to the point cloud semantic segmentation result, perform clustering processing on the point cloud data of the transmission tower, and obtain point cloud data sets of multiple different transmission towers.

[0084] It should be noted that, since the point cloud segmentation processing is performed above, the segmentation processing may easily cause a segment of point cloud data to contain 0 to N point clouds and 0 to 2 incomplete point clouds.

[0085] Therefore, it is necessary to cluster the point cloud data of the transmission tower, and the clustering method can adopt the DBSCAN clustering method.

[0086] Step S4: perform point cloud division according to the point cloud data set of the transmission tower, divide the point cloud data of the transmission line associated with the transmission tower, and update the point cloud data of the transmission line to the point cloud data set of the transmission tower.

[0087] Among them, since the point cloud data set of the transmission tower in the aforementioned step S3 only contains tower points, in practical applications, such as in the lightning protection analysis of transmission lines, it is necessary to extract information such as tower geometric parameters and conductor hanging points, so it is necessary to re-divide the point cloud data set of the transmission tower.

[0088] Specifically, if Figure 3 As shown, step S4 specifically includes:

[0089] Step S401: perform mean operation on the point cloud data set of the transmission tower to obtain the centroid point cloud coordinates of the transmission tower, and determine the point cloud centroid distance between each point cloud data in the point cloud data set of the transmission tower and the centroid point cloud coordinates.

[0090] Step S402: Filter out a plurality of point cloud data with a point cloud centroid distance less than a preset distance threshold, and construct a tower centroid point cloud data set.

[0091] The distance threshold may be ΔL / 2.

[0092] Step S403: determine a second minimum enclosing rectangle based on the tower gravity center point cloud data set.

[0093] Step S404: determine the tower width according to the shortest side of the second minimum enclosing rectangle.

[0094] Step S405: determine a circular area with the centroid point cloud coordinates of the transmission tower as the origin and the tower width at a preset multiple as the radius, and determine all point cloud data within the circular area as point cloud data of the transmission line associated with the transmission tower.

[0095] Among them, the point cloud data of the transmission line includes conductor points and insulator string points.

[0096] Step S406: Update the point cloud data of the transmission line associated with the transmission tower into the point cloud data set of the transmission tower.

[0097] Step S5: extracting key point cloud parameters of the transmission tower according to the updated point cloud data set of the transmission tower.

[0098] It should be noted that there are many types of transmission line towers with various structural forms. There are large differences in the geometric features of the towers, the combination of insulator strings, and the location of the conductor hanging points. It is difficult to use a universal method to extract and analyze the key parameters of all types of tower point clouds. Therefore, it is necessary to first design a tower type recognition algorithm that can identify the types of towers, such as dry towers, cat head towers, and wine glass towers, and then process the tower point clouds of each different tower type separately. The tower type is classified using a point cloud classification method based on deep learning.

[0099] In a specific embodiment, Figure 4 As shown, step S5 specifically includes:

[0100] Step S501: Project the updated point cloud data set of the transmission tower onto a two-dimensional coordinate plane to obtain a two-dimensional coordinate point set.

[0101] Step S502: Process the two-dimensional coordinate point set based on the PCA algorithm to obtain the main plane vector direction.

[0102] The PCA (principle component analysis) algorithm is the principal component analysis method. By inputting a two-dimensional coordinate point set into the PCA algorithm, the main plane vector direction can be obtained. The main plane vector direction includes the transverse vector direction of the tower main plane projection and the vertical vector direction of the tower main plane projection. Among them, the vector direction with the largest degree is the transverse vector direction of the tower main plane projection, and the other direction is the vertical vector direction of the tower main plane projection.

[0103] In practical applications, the direction of the main plane vector may exist in two directions, so it is necessary to identify the direction of the main plane vector. Specifically, it includes:

[0104] Step S5021: determine whether the angle between the vector direction between the center point of the transmission tower and the center point of its adjacent transmission tower and the horizontal vector direction of the tower main plane projection is an acute angle.

[0105] Step S5022: If it is determined that the angle is not an acute angle, the horizontal vector direction of the main plane projection of the tower is rotated to 180° along the Z axis with the center of gravity of the tower as the origin, and the horizontal vector direction of the main plane projection of the tower obtained after the rotation is updated to the new horizontal vector direction of the main plane projection of the tower.

[0106] It should be noted that since the two vectors obtained by the PCA algorithm have no directionality, the main plane to be obtained must be along the direction of the transmission line (the vector direction between the center point of the transmission tower and the center point of its adjacent transmission tower). The vertical vector direction of the tower main plane projection and the direction of the main plane along the transmission line should be an acute angle.

[0107] Step S503: Rotate and project the point cloud data set of the transmission tower into a grid in the direction of the main plane vector to generate a two-dimensional image.

[0108] Step S504: perform edge smoothing processing on the two-dimensional image.

[0109] Among them, the Gaussian blur method can be used to perform edge smoothing on the two-dimensional image.

[0110] Step S505: Use the Canny edge detection operator to perform edge detection on the two-dimensional image after edge smoothing, and extract the maximum outer contour according to the edge detection result.

[0111] Among them, since the tower corners often have obstruction problems, resulting in edge deformities, the tower corners need to be truncated, and the truncation can be started from the lowest crossarm position.

[0112] Step S506: Use the LSD straight line detection algorithm to perform straight line segment detection on the maximum outer contour, and determine multiple straight line segments whose lengths are less than a preset length threshold according to the straight line segment detection result.

[0113] Among them, the LSD (Linear Discriminant Analysis) line detection algorithm is a classic two-classification algorithm.

[0114] Step S507: merge the selected multiple straight line segments, and extract the intersection points between the merged multiple straight line segments as the key corner points of the transmission tower.

[0115] Among them, the intersection of the above straight line segments is the key corner point of the transmission tower, and the intersection coordinate parameters are the coordinates of the key corner point of the transmission tower. However, since the point cloud data set of the transmission tower is rotated and projected into the grid in the direction of the main plane vector, the coordinates of the key corner point do not include the y-axis coordinate.

[0116] To this end, the embodiment of the present invention further includes:

[0117] Step S5071, determining the XZ two-dimensional coordinates of the key corner points of the transmission tower according to the intersection points;

[0118] Step S5072: Project the point cloud data set of the transmission tower onto a two-dimensional plane in the direction of the side view vector, and determine the maximum coordinate value in the Y-axis direction corresponding to the Z-axis coordinate value of the key corner point of the transmission tower based on the projection of the transmission tower on the two-dimensional plane in the direction of the side view vector.

[0119] The side view projection of the tower can be approximated as a trapezoid that is narrow at the top and wide at the bottom, such as Figure 5 As shown. The projection of the tower corner point in the side view is located on the two hypotenuses of the trapezoid. Therefore, according to the z-axis coordinate value of the two-dimensional corner point coordinate, the maximum value of the left and right sides of the side view relative to the z-axis coordinate value can be taken as the y-axis coordinate of the key corner point.

[0120] Step S5073: determine the three-dimensional coordinates of the key corner points of the transmission tower according to the maximum coordinate value in the Y-axis direction and the XZ two-dimensional coordinates of the key corner points of the transmission tower.

[0121] In a specific embodiment, the method further includes conductor catenary equation fitting and insulator string fitting.

[0122] like Figure 6 As shown, the wire catenary equation fitting includes:

[0123] Step S5081: According to the point cloud semantic segmentation result, the point cloud data of the conductor is rotated to the main plane vector direction, an O-XYZ three-dimensional coordinate system is established with the center of gravity of the transmission tower as the origin, and the point cloud data of the conductor is divided into point cloud data of the conductors on both sides through the XOZ plane;

[0124] Step S5082, clustering the point cloud data of the conductors on both sides respectively to obtain multiple single-phase conductors;

[0125] Step S5083, determining a coordinate sequence data set of each single-phase conductor according to the coordinate position distribution of each single-phase conductor in the O-XYZ three-dimensional coordinate system;

[0126] Step S5084: Use the RANSAC plane fitting algorithm to fit the coordinate sequence data set of the single-phase conductor to obtain the plane equation of the single-phase conductor;

[0127] Step S5085: Use the RANSAC line fitting algorithm to fit the coordinate sequence data set of the single-phase conductor to obtain the catenary equation of the single-phase conductor.

[0128] In the insulator string fitting, the insulator string can be approximated as a straight line. The RANSAC straight line fitting algorithm is used to fit the point cloud data of the insulator string to obtain the insulator string fitting equation.

[0129] In addition, in the lightning protection analysis, GIS data, front side span and line angle etc. are also needed, so the present invention also includes:

[0130] Extract GIS data, front span and line turning angle based on the updated point cloud data set of transmission towers.

[0131] Among them, for GIS data extraction, a lightning density and terrain database is established based on the longitude and latitude coordinates. Through the coordinates of the center of gravity of the point cloud, key parameters such as lightning density, terrain data, and ground inclination are extracted from the GIS database.

[0132] For the extraction of front side span and line angle. The front side span is the tower T i-1 To tower T i The horizontal distance between.

[0133] The line corner is the tower T i-1 To tower T i Vector With tower T i To tower T i+1 Vector The angle between Since the value of θ calculated by this formula is a positive value, the angle direction between the two vectors must also be considered. when When the result is positive, the line

[0135] The road turning angle θ`=θ, otherwise, the road turning angle θ`=-θ.

[0136] It should be noted that the present invention performs point cloud semantic segmentation on point cloud data to determine point cloud data corresponding to multiple different device types, clusters the point cloud data of the transmission tower, divides the point cloud according to the point cloud data set of the transmission tower obtained by clustering, divides the point cloud data of the transmission line associated with the transmission tower, and updates it to the point cloud data set of the transmission tower, thereby accurately covering the point cloud data of the transmission line and extracting key point cloud parameters of the transmission tower, thereby avoiding the problems of lost information, redundant and complex calculation content, and excessive memory usage in the calculation process, and improving the accuracy and efficiency of point cloud parameter extraction.

[0137] The above is a detailed description of an embodiment of a method for extracting point cloud parameters of a transmission line provided by the present invention. The following is a detailed description of an embodiment of a system for extracting point cloud parameters of a transmission line provided by the present invention.

[0138] like Figure 7 As shown, Figure 7 The structure of a transmission line point cloud parameter extraction system provided by the present invention is illustrated.

[0139] The present invention provides a transmission line point cloud parameter extraction system, comprising:

[0140] The preprocessing module 100 is used to preprocess the point cloud data of the transmission line.

[0141] Among them, the point cloud data of the transmission line is obtained by scanning along the transmission line with a drone carrying an airborne lidar, which includes point sets such as transmission lines, towers, insulators, and nearby ground, trees, and buildings.

[0142] Among them, the preprocessing includes point cloud segmentation processing and point cloud denoising processing.

[0143] The semantic segmentation module 200 is used to perform point cloud semantic segmentation on the preprocessed point cloud data. The point cloud semantic segmentation is used to determine the point cloud data corresponding to multiple different device types.

[0144] Among them, a semantic segmentation model based on deep learning can be used to perform semantic segmentation on point cloud data. Point cloud semantic segmentation can determine the point cloud data corresponding to multiple different equipment types. The equipment types include but are not limited to insulator string points, ground points, transmission line points and transmission tower points.

[0145] Since there are some misclassifications in the semantic segmentation results, the point cloud data needs to be denoised and filtered again. For the semantically segmented point cloud, the point cloud data is clustered according to the semantic segmentation labels to remove the misclassified data points.

[0146] Exemplarily, the DBSCAN clustering method can be used to perform DBSCAN clustering on each type of point set, and points with a number of points below a threshold are regarded as noise, thereby filtering out noise points and removing misclassified data points.

[0147] The point cloud clustering module 300 is used to determine the point cloud data of the transmission tower according to the point cloud semantic segmentation result, and perform clustering processing on the point cloud data of the transmission tower to obtain a plurality of different point cloud data sets of the transmission tower.

[0148] It should be noted that, since the point cloud segmentation processing is performed above, the segmentation processing may easily cause a segment of point cloud data to contain 0 to N point clouds and 0 to 2 incomplete point clouds.

[0149] Therefore, it is necessary to cluster the point cloud data of the transmission tower, and the clustering method can adopt the DBSCAN clustering method.

[0150] The point cloud division module 400 is used to perform point cloud division according to the point cloud data set of the transmission tower, divide the point cloud data of the transmission line associated with the transmission tower, and update the point cloud data of the transmission line to the point cloud data set of the transmission tower.

[0151] Among them, since the point cloud data set of the transmission tower in the aforementioned step S3 only contains tower points, in practical applications, such as in the lightning protection analysis of transmission lines, it is necessary to extract information such as tower geometric parameters and conductor hanging points, so it is necessary to re-divide the point cloud data set of the transmission tower.

[0152] The point cloud parameter extraction module 500 is used to extract key point cloud parameters of the transmission tower according to the updated point cloud data set of the transmission tower.

[0153] Specifically, the updated point cloud data set of the transmission tower is projected onto a two-dimensional coordinate plane to obtain a two-dimensional coordinate point set.

[0154] The two-dimensional coordinate point set is processed based on the PCA algorithm to obtain the main plane vector direction.

[0155] The PCA (principle component analysis) algorithm is the principal component analysis method. By inputting a two-dimensional coordinate point set into the PCA algorithm, the main plane vector direction can be obtained. The main plane vector direction includes the transverse vector direction of the tower main plane projection and the vertical vector direction of the tower main plane projection. Among them, the vector direction with the largest degree is the transverse vector direction of the tower main plane projection, and the other direction is the vertical vector direction of the tower main plane projection.

[0156] In practical applications, the direction of the main plane vector may exist in two directions, so it is necessary to identify the direction of the main plane vector. Specifically, it includes:

[0157] Determine whether the angle between the vector direction of the center point of the transmission tower and the center point of its adjacent transmission tower and the horizontal vector direction of the tower main plane projection is an acute angle.

[0158] If the angle is not an acute angle, the transverse vector direction of the tower main plane projection is rotated to 180° along the Z axis with the center of gravity of the tower as the origin, and the transverse vector direction of the tower main plane projection obtained after the rotation is updated to the new transverse vector direction of the tower main plane projection.

[0159] It should be noted that since the two vectors obtained by the PCA algorithm have no directionality, the main plane to be obtained must be along the direction of the transmission line (the vector direction between the center point of the transmission tower and the center point of its adjacent transmission tower). The vertical vector direction of the tower main plane projection and the direction of the main plane along the transmission line should be an acute angle.

[0160] The point cloud data set of the transmission tower is rotated and projected into a grid in the direction of the principal plane vector to generate a two-dimensional image.

[0161] Perform edge smoothing on a 2D image.

[0162] Among them, the Gaussian blur method can be used to perform edge smoothing on the two-dimensional image.

[0163] The Canny edge detection operator is used to perform edge detection on the two-dimensional image after edge smoothing, and the maximum outer contour is extracted according to the edge detection result.

[0164] Among them, since the tower corners often have obstruction problems, resulting in edge deformities, the tower corners need to be truncated, and the truncation can be started from the lowest crossarm position.

[0165] The LSD straight line detection algorithm is used to perform straight line segment detection on the maximum outer contour, and multiple straight line segments whose lengths are less than a preset length threshold are determined according to the straight line segment detection results.

[0166] Among them, the LSD (Linear Discriminant Analysis) line detection algorithm is a classic two-classification algorithm.

[0167] The selected multiple straight line segments are merged, and the intersection points between the merged multiple straight line segments are extracted as the key corner points of the transmission tower.

[0168] Among them, the intersection of the above straight line segments is the key corner point of the transmission tower, and the intersection coordinate parameters are the coordinates of the key corner point of the transmission tower. However, since the point cloud data set of the transmission tower is rotated and projected into the grid in the direction of the main plane vector, the coordinates of the key corner point do not include the y-axis coordinate.

[0169] To this end, the embodiment of the present invention further includes:

[0170] Determine the XZ two-dimensional coordinates of the key corner points of the transmission tower based on the intersection points;

[0171] The point cloud data set of the transmission tower is projected onto a two-dimensional plane in the direction of the side view vector. Based on the projection of the transmission tower on the two-dimensional plane in the direction of the side view vector, the maximum coordinate value in the Y-axis direction corresponding to the Z-axis coordinate value of the key corner point of the transmission tower is determined.

[0172] The side view projection of the tower can be approximated as a trapezoid that is narrow at the top and wide at the bottom, such as Figure 5 As shown. The projection of the tower corner point in the side view is located on the two hypotenuses of the trapezoid. Therefore, according to the z-axis coordinate value of the two-dimensional corner point coordinate, the maximum value of the left and right sides of the side view relative to the z-axis coordinate value can be taken as the y-axis coordinate of the key corner point.

[0173] Step S5073: determine the three-dimensional coordinates of the key corner points of the transmission tower according to the maximum coordinate value in the Y-axis direction and the XZ two-dimensional coordinates of the key corner points of the transmission tower.

[0174] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working process of the system described above can refer to the corresponding process in the aforementioned method embodiment, and will not be repeated here.

[0175] In several embodiments provided by the present invention, it is understood that each box in the flow chart or block diagram can represent a module, a program segment or a part of a code, and a part of a module, a program segment or a code includes one or more executable instructions for realizing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in a different order from that marked in the accompanying drawings. For example, two consecutive boxes can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved.

[0176] In the several embodiments provided by the present invention, it should be understood that the disclosed systems and methods can be implemented in other ways. For example, the device embodiments described above are only schematic, for example, the division of units is only a logical function division, and there may be other division methods in actual implementation, such as multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed. Another point is that the mutual coupling or direct coupling or communication connection shown or discussed can be through some interfaces, indirect coupling or communication connection of devices or units, which can be electrical, mechanical or other forms.

[0177] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features thereof may be replaced by equivalents. However, these modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention.

Claims

1. A method for extracting point cloud parameters of transmission lines, It is characterized in that include: Preprocessing of point cloud data of transmission lines; Performing point cloud semantic segmentation on the preprocessed point cloud data, wherein the point cloud semantic segmentation is used to determine point cloud data corresponding to a plurality of different device types; Determine the point cloud data of the transmission tower according to the point cloud semantic segmentation result, perform clustering processing on the point cloud data of the transmission tower, and obtain a plurality of different point cloud data sets of the transmission tower; Performing point cloud division according to the point cloud data set of the transmission tower, dividing the point cloud data of the transmission line associated with the transmission tower, and updating the point cloud data of the transmission line to the point cloud data set of the transmission tower; The key point cloud parameters of the transmission tower are extracted according to the updated point cloud data set of the transmission tower.

2. The method for extracting point cloud parameters of a transmission line according to claim 1, It is characterized in that The preprocessing includes point cloud segmentation processing and point cloud denoising processing.

3. The method for extracting point cloud parameters of a transmission line according to claim 2, It is characterized in that The step of point cloud segmentation processing includes: Determining boundary point cloud data according to the point cloud data of the transmission line; Determine, according to the boundary point cloud data, a first minimum enclosing rectangle of the point cloud data of the transmission line as a boundary box of the transmission line; Point cloud segmentation is performed according to the maximum side length of the boundary box of the power transmission line according to a preset segment moving step length, and the point cloud data of the power transmission line is segmented into point cloud data of a plurality of segments.

4. The method for extracting point cloud parameters of a transmission line according to claim 1, It is characterized in that The step of dividing the point cloud according to the point cloud data set of the transmission tower, dividing the point cloud data of the transmission line associated with the transmission tower, and updating the point cloud data of the transmission line to the point cloud data set of the transmission tower specifically includes: Performing a mean operation on the point cloud data set of the transmission tower to obtain the centroid point cloud coordinates of the transmission tower, and determining the point cloud centroid distance between each point cloud data in the point cloud data set of the transmission tower and the centroid point cloud coordinates; Filter out a plurality of point cloud data whose point cloud gravity center distance is less than a preset distance threshold, and construct a tower gravity center point cloud data set; Determine a second minimum enclosing rectangle based on the tower gravity center point cloud data set; Determine the width of the tower according to the shortest side of the second minimum enclosing rectangle; Taking the point cloud coordinates of the center of gravity of the transmission tower as the origin and the width of the tower at a preset multiple as the radius, a circular area is determined, and all point cloud data within the circular area are determined to be point cloud data of the transmission line associated with the transmission tower; The point cloud data of the transmission line associated with the transmission tower is updated into the point cloud data set of the transmission tower.

5. The method for extracting point cloud parameters of a transmission line according to claim 1, It is characterized in that The step of extracting key point cloud parameters of the transmission tower according to the updated point cloud data set of the transmission tower specifically includes: Projecting the updated point cloud data set of the transmission tower onto a two-dimensional coordinate plane to obtain a two-dimensional coordinate point set; Processing the two-dimensional coordinate point set based on the PCA algorithm to obtain the main plane vector direction; Rotating and projecting the point cloud data set of the transmission tower into a grid in the direction of the main plane vector to generate a two-dimensional image; Performing edge smoothing processing on the two-dimensional image; The Canny edge detection operator is used to perform edge detection on the two-dimensional image after edge smoothing, and the maximum outer contour is extracted according to the edge detection result; Using an LSD straight line detection algorithm to perform straight line segment detection on the maximum outer contour, and determining a plurality of straight line segments whose lengths are less than a preset length threshold according to the straight line segment detection results; The screened multiple straight line segments are merged, and the intersection points between the merged multiple straight line segments are extracted as the key corner points of the transmission tower.

6. The method for extracting point cloud parameters of a transmission line according to claim 5, It is characterized in that The main plane vector direction includes the transverse vector direction of the tower main plane projection and the vertical vector direction of the tower main plane projection; After the step of processing the two-dimensional coordinate point set based on the PCA algorithm to obtain the main plane vector direction, the method further includes: Determine whether the angle between the vector direction between the center point of the transmission tower and the center point of its adjacent transmission tower and the horizontal vector direction of the main plane projection of the tower is an acute angle; If it is determined that the angle is not an acute angle, the transverse vector direction of the projection of the main plane of the tower is rotated to 180° along the Z-axis direction with the center of gravity of the tower as the origin, and the transverse vector direction of the projection of the main plane of the tower obtained after the rotation is updated to the new transverse vector direction of the projection of the main plane of the tower.

7. The method for extracting point cloud parameters of a transmission line according to claim 5, It is characterized in that After the step of merging the selected multiple straight line segments and extracting the intersection points between the merged multiple straight line segments as the key corner points of the transmission tower, the method further includes: Determine the XZ two-dimensional coordinates of the key corner points of the transmission tower according to the intersection points; Projecting the point cloud data set of the transmission tower onto a two-dimensional plane in the direction of the side view vector, and determining the maximum coordinate value in the Y-axis direction corresponding to the Z-axis coordinate value of the key corner point of the transmission tower based on the projection of the transmission tower onto the two-dimensional plane in the direction of the side view vector; The three-dimensional coordinates of the key corner points of the transmission tower are determined according to the maximum coordinate value in the Y-axis direction and the XZ two-dimensional coordinates of the key corner points of the transmission tower.

8. The method for extracting point cloud parameters of a transmission line according to claim 5, It is characterized in that After the step of extracting the key point cloud parameters of the transmission tower according to the updated point cloud data set of the transmission tower, the method further includes: According to the point cloud semantic segmentation result, the point cloud data of the conductor is rotated to the main plane vector direction, an O-XYZ three-dimensional coordinate system is established with the center of gravity of the transmission tower as the origin, and the point cloud data of the conductor is divided into point cloud data of the conductors on both sides through the XOZ plane; The point cloud data of the conductors on both sides are clustered to obtain multiple single-phase conductors; Determine a coordinate sequence data set of each single-phase conductor according to the coordinate position distribution of each single-phase conductor in the O-XYZ three-dimensional coordinate system; Using the RANSAC plane fitting algorithm to fit the coordinate sequence data set of the single-phase conductor to obtain the plane equation of the single-phase conductor; The RANSAC line fitting algorithm is used to fit the coordinate sequence data set of the single-phase conductor to obtain the catenary equation of the single-phase conductor.

9. The method for extracting point cloud parameters of a transmission line according to claim 1, It is characterized in that Also includes: Extract GIS data, front span and line turning angle based on the updated point cloud data set of transmission towers.

10. A transmission line point cloud parameter extraction system, It is characterized in that include: A preprocessing module, used for preprocessing point cloud data of transmission lines; A semantic segmentation module, used to perform point cloud semantic segmentation on the preprocessed point cloud data, wherein the point cloud semantic segmentation is used to determine point cloud data corresponding to a plurality of different device types; A point cloud clustering module is used to determine the point cloud data of the transmission tower according to the point cloud semantic segmentation result, and perform clustering processing on the point cloud data of the transmission tower to obtain a plurality of different point cloud data sets of the transmission tower; A point cloud division module, used to perform point cloud division according to the point cloud data set of the transmission tower, divide the point cloud data of the transmission line associated with the transmission tower, and update the point cloud data of the transmission line to the point cloud data set of the transmission tower; The point cloud parameter extraction module is used to extract key point cloud parameters of the transmission tower according to the updated point cloud data set of the transmission tower.

Images