A method for automatic separation of power stages without auxiliary data based on LiDAR point cloud

Through digital elevation correlation model and mathematical morphology processing, the unorganized lidar point cloud data is used to realize the automatic separation of power corridor segments, solving the problems of artificial dependence and low computing efficiency in the existing technology, and improving power inspection degree of automation and intelligence.

CN118570453BActive Publication Date: 2024-11-08JIANGMEN POWER SUPPLY BUREAU OF GUANGDONG POWER GRID CO LTD
View PDF 4 Cites 0 Cited by

Patent Information

Application Number
CN202411053007.X
Authority / Receiving Office
CN · China
Patent Type
Patents(China)
Current Assignee / Owner
Filing Date
2024-08-02
Publication Date
2024-11-08
Estimated Expiration
2044-08-02

AI Technical Summary

Technical Problem

In the prior art, the automatic separation algorithm for power corridor segments relies on manual visual interpretation and prior information, has low computing efficiency, high resource utilization, and lacks effective automated positioning technology for unorganized lidar point cloud data.

Method used

The difference calculation elevation feature map is used to use digital elevation correlation models (DEM, DSM) to perform the difference calculation elevation feature map. Through the height difference feature and the threshold of the wingspan length of the power pole tower, automatic separation of the power segment without auxiliary lidar point cloud data is achieved, combined with mathematical morphology processing and iterate the target search, remove non-power elements and linear structures, and calculate the minimum external rectangle and center point of the power tower.

Benefits of technology

It improves the positioning accuracy and characteristic calculation efficiency of power pole towers, realizes the automatic separation of power corridor sections, improves the automation and intelligence of power inspections, reduces the occupation of computing resources, and simplifies the algorithm process.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN118570453B_ABST
    Figure CN118570453B_ABST
Patent Text Reader

Abstract

The present invention provides a method for automatic separation of power sections without auxiliary data based on laser radar point cloud, comprising the following steps: preprocessing laser radar point cloud; grid division and projection; obtaining height difference feature map; removing non-power element height difference significant response area; obtaining pixel candidate area of ​​power pole tower and power line; removing linear structure pixels; iterative target contour search; convex hulling to calculate the minimum circumscribed rectangle; calculating the center point of the circumscribed rectangle to locate the power pole tower; determining the rectangular frame area along the line; extracting laser point cloud data in the area; the algorithm stops when all the projection coordinates of the power pole tower project are taken out. The algorithm is aimed at the problems of insufficient automation of power inspection, high computing resource occupancy rate, low application computing efficiency, etc., and realizes automation and intelligent separation of power sections based on unorganized laser radar point cloud, so as to improve the automation and intelligence of power inspection and improve the efficiency of intelligent analysis of subsequent applications of power inspection.
Need to check novelty before this filing date? Find Prior Art

Description

Technical Field

[0001] The present invention relates to the technical field of electric power inspection, and in particular to a method for automatic separation of electric power stages without auxiliary data based on laser radar point cloud. Background Art

[0002] In recent years, the characteristics of my country's power industry, such as large scale, long chain, wide coverage and strong social dependence, have become increasingly prominent. The traditional manual power inspection technology that consumes a lot of manpower and time can no longer meet the requirements of the industry. The demand for new automated, digital and intelligent power corridor inspection systems with high-performance computing and massive information data analysis capabilities is increasing.

[0003] The laser radar point cloud scanning data in the power corridor scene is affected by factors such as the laser radar measurement height and scanning angle. It usually contains a large number of non-focus points and has a large data volume and high memory resource occupancy rate. The automatic separation algorithm of power corridor sections can automatically and effectively lock and separate the focus areas along the power section to reduce the computer memory occupancy rate and improve the computing efficiency of subsequent applications. In terms of automatic positioning of power towers based on unorganized laser point cloud data in power corridor scenes, there are only a few studies on automatic positioning of transmission towers from laser point clouds at home and abroad. Among them, the separation of power corridor sections mostly relies on manual visual interpretation and prior power tower location information.

[0004] The prior art has not yet provided relevant content on the algorithm for automatic separation of power corridor sections, but the positioning of power poles and towers based on unorganized lidar point clouds is the key to automatic separation of power corridor sections. The existing power pole and tower positioning technology based on unorganized lidar point clouds has the following disadvantages: 1. Most of them rely on manual visual interpretation and processing and prior power pole and tower position information, and the degree of automation of power inspection is low; 2. The automatic positioning of power poles and towers requires a large number of features and requires the assistance of overhead power lines, so the calculation efficiency is low and more computing resources are occupied; 3. The algorithm is relatively complex and there are multiple empirical thresholds to be determined. Summary of the invention

[0005] In view of the shortcomings of the prior art, the purpose of the present invention is to provide a method for automatic separation of power sections without auxiliary data based on laser radar point cloud, so as to solve the problems raised in the above-mentioned background technology. The present invention selects efficient and significant digital elevation related models (DEM, DSM) to perform differential calculation of elevation feature maps, which can not only locate power poles and towers with high accuracy, but also improve the efficiency of feature calculation. Only two thresholds of height difference features and power pole tower wingspan length are set and only the input of unorganized laser point cloud data is required, so as to realize relatively robust automatic separation of power corridor sections.

[0006] In order to achieve the above object, the present invention is implemented by the following technical solution: a method for automatic separation of power stages without auxiliary data based on laser radar point cloud, comprising the following steps:

[0007] S1, performing outlier statistical filtering on the original acquired lidar point cloud to remove outliers;

[0008] S2, gridding and projecting the preprocessed laser point cloud;

[0009] S3, constructing the digital elevation feature map of the quasi-DEM and quasi-DSM statistically and making a difference to obtain the height difference feature map;

[0010] S4. Remove the area with significant response to the height difference of non-power elements;

[0011] S5, obtaining pixel candidate regions of power poles and power lines;

[0012] S6, removing linear structure pixels;

[0013] S7, iterative target contour search, stop when the number of target contours changes from before iteration;

[0014] S8, performing convex hull transformation based on the extracted contour to calculate the minimum enclosing rectangle;

[0015] S9, calculating the center point of the circumscribed rectangle to locate the power tower and inversely calculating the engineering projection coordinates based on the image coordinates;

[0016] S10, sequentially taking out the projection coordinates of two adjacent power tower projects and determining the rectangular frame area along the line based on the prior span width of the power tower;

[0017] S11, extracting laser point cloud data according to the rectangular frame area along the power stage;

[0018] S12. When all the projection coordinates of the power tower project are taken out, the algorithm stops.

[0019] Furthermore, the grid size of the grid division in S2 is 1m×1m.

[0020] Furthermore, the removal process in S4 performs coarse removal based on median filtering.

[0021] Furthermore, in S5, a threshold is set according to actual experience and a feature is generated based on a cumulative histogram of height difference features. Figure 2 Value.

[0022] Furthermore, the removal process of linear structure pixels needs to meet the requirements of the high-voltage overhead transmission line device, and a 5×5 sliding window is set based on the morphological opening operation to complete the removal process.

[0023] Furthermore, after the laser point cloud data is filtered, all point clouds within the rectangular frame area are retained.

[0024] Furthermore, all point clouds within the rectangular frame area are generated into independent storage files in .txt format.

[0025] Beneficial effects of the present invention:

[0026] 1. This method for automatic separation of power sections without auxiliary data based on lidar point cloud can not only locate power poles and towers with high accuracy, but also improve the efficiency of feature calculation by selecting efficient and significant digital elevation related models (DEM, DSM) to calculate the elevation feature map by difference. It only sets two thresholds of height difference feature and power pole span length and only requires the input of unorganized laser point cloud data to achieve relatively robust automatic separation of power corridor sections.

[0027] 2. The present invention aims to solve the problems of insufficient automation of power inspection, high occupancy rate of computing resources, low application computing efficiency, etc., and realizes the automation and intelligent separation of power sections based on unorganized lidar point cloud, so as to improve the automation and intelligence of power inspection and enhance the efficiency of intelligent analysis of subsequent applications of power inspection. At the same time, it also improves the automatic positioning algorithm of power towers and extends it to realize the application of automatic separation of power corridor sections.

[0028] 3. The present invention adopts more streamlined and effective height difference features, avoids the input of redundant features and non-significant features, improves the feature calculation efficiency, and speeds up the algorithm running time; introduces a mathematical morphology processing method to effectively remove most of the height difference significant pixels of non-power towers, and uses an iterative shrinkage method to approximate the plane center position of the power tower; expands subsequent applications, extends the automatic positioning of power towers to the automatic separation application of power corridor sections, and enhances the degree of automation of power inspections. BRIEF DESCRIPTION OF THE DRAWINGS

[0029] Figure 1 This is a flow chart of a method for automatic separation of power stages without auxiliary data based on laser radar point cloud according to the present invention;

[0030] Figure 2 The 1m×1m spatial resolution pseudo-DEM, pseudo-DSM and height difference feature display map of the present invention;

[0031] Figure 3 This is a diagram showing the center point positioning of the power tower of the present invention;

[0032] Figure 4 This is a result diagram of automatic separation of power corridor sections based on laser radar point cloud of the present invention;

[0033] Figure 5 It is a flow chart of the separation algorithm in the prior art. DETAILED DESCRIPTION

[0034] In order to make the technical means, creative features, objectives and effects achieved by the present invention easy to understand, the present invention is further explained below in conjunction with specific implementation methods.

[0035] See also Figures 1 to 5 ,in Figure 5 The flowchart of the separation algorithm in the prior art, the present invention provides the following technical solutions: a method for automatic separation of power sections without auxiliary data based on laser radar point cloud, the automatic separation algorithm can locate power poles and towers with high accuracy and improve the efficiency of feature calculation by selecting efficient and significant digital elevation related models (DEM, DSM) to calculate the elevation feature map by difference; only two thresholds of height difference feature and power pole span length are set and only the input of unorganized laser point cloud data is required to achieve relatively robust automatic separation of power corridor sections. As shown in the attached figure Figure 1 The flowchart of the method for automatic separation of power stages without auxiliary data based on laser radar point cloud is shown, wherein the specific execution process of the algorithm is as follows:

[0036] 1. Perform outlier statistical filtering on the original acquired LiDAR point cloud to remove outliers;

[0037] 2. Divide the preprocessed laser point cloud into a grid of 1m×1m and project it;

[0038] 3. Statistically construct the digital elevation feature map of the quasi-DEM and quasi-DSM and make a difference to obtain the height difference feature map; Figure 2 As shown, this embodiment provides a pseudo-DEM, pseudo-DSM and height difference feature display diagram based on the above 1m×1m spatial resolution;

[0039] 4. Use median filtering to roughly remove the areas with significant response to non-power element height differences;

[0040] 5. Set the threshold according to actual experience and perform feature extraction based on the cumulative histogram of height difference features. Figure 2 Value, obtain pixel candidate areas of power towers and power lines;

[0041] 6. According to the relevant provisions of the high-voltage overhead transmission line device, a 5×5 sliding window is set based on the morphological opening operation to remove the linear structure pixels. For the process of removing non-power element pixels in the height difference feature map, it can also be achieved by selecting a variety of conventional digital image processing methods according to the relevant national provisions of the high-voltage overhead transmission line device;

[0042] 7. Iterate the target contour search and stop when the number of target contours changes from before the iteration;

[0043] 8. Perform convex hull transformation based on the extracted contour to calculate the minimum enclosing rectangle;

[0044] 9. Calculate the center point of the circumscribed rectangle to locate the power tower and inversely calculate the engineering projection coordinates based on the image coordinates; wherein, the power tower center point positioning provided in this embodiment is shown as follows Figure 3 As shown;

[0045] 10. Take out the projection coordinates of two adjacent power tower projects in turn and determine the rectangular frame area along the line based on the prior span width of the power tower;

[0046] 11. Extract laser point cloud data based on the rectangular frame area along the power section, retain all point clouds in the rectangular frame area and generate independent storage files in .txt format. And provide automatic separation results of power corridor sections based on laser radar point cloud. The separation results are as follows: Figure 4 shown.

[0047] 12. When all the projection coordinates of the power tower project are taken out, the algorithm stops.

[0048] The present invention adopts more streamlined and effective height difference features, avoids the input of redundant features and non-significant features, improves the feature calculation efficiency, and speeds up the algorithm running time; introduces a mathematical morphology processing method to effectively remove most of the height difference significant pixels of non-power towers, and uses an iterative contraction method to approximate the plane center position of the power tower; expands subsequent applications, extends the automatic positioning of power towers to the automatic separation of power corridor sections, and enhances the automation of power inspection. At the same time, the automatic positioning algorithm of power towers is also improved, and the application of automatic separation of power corridor sections is realized.

[0049] The basic principles and main features of the present invention and the advantages of the present invention are shown and described above. For those skilled in the art, it is obvious that the present invention is not limited to the details of the above exemplary embodiments, and the present invention can be implemented in other specific forms without departing from the spirit or basic characteristics of the present invention.

[0050] In addition, it should be understood that although the present specification is described according to implementation modes, not every implementation mode contains only one independent technical solution. This description of the specification is only for the sake of clarity. Those skilled in the art should regard the specification as a whole. The technical solutions in each embodiment may also be appropriately combined to form other implementation modes that can be understood by those skilled in the art.

Claims

1. A method for automatic separation of power stages without auxiliary data based on laser radar point cloud, characterized in that: The following steps are involved: S1, performing outlier statistical filtering on the original acquired lidar point cloud to remove outliers; S2, gridding and projecting the preprocessed laser point cloud; S3, constructing the digital elevation feature map of the quasi-DEM and quasi-DSM statistically and making a difference to obtain the height difference feature map; S4. Remove the area with significant response to the height difference of non-power elements; S5, obtaining pixel candidate regions of power poles and power lines; S6, removing linear structure pixels; S7, iterative target contour search, stop when the number of target contours changes from before iteration; S8, performing convex hull transformation based on the extracted contour to calculate the minimum enclosing rectangle; S9, calculating the center point of the circumscribed rectangle to locate the power tower and inversely calculating the engineering projection coordinates based on the image coordinates, where the engineering projection coordinates are UTM area projections; S10, sequentially taking out the projection coordinates of two adjacent power tower projects and determining the rectangular frame area along the line based on the prior span width of the power tower; S11, extracting laser point cloud data according to the rectangular frame area along the power stage; S12. When all the projection coordinates of the power tower project are taken out, the algorithm stops.

2. According to claim 1, a method for automatic separation of power stages without auxiliary data based on laser radar point cloud, characterized in that: The grid size of the grid division in S2 is 1m×1m.

3. The method for automatic separation of power stages without auxiliary data based on laser radar point cloud according to claim 1 is characterized in that: The removal process in S4 is based on median filtering to perform rough removal.

4. The method for automatic separation of power stages without auxiliary data based on laser radar point cloud according to claim 1 is characterized in that: In S5, a threshold is set according to actual experience and the feature map is binarized based on the height difference feature cumulative histogram.

5. The method for automatic separation of power stages without auxiliary data based on laser radar point cloud according to claim 1 is characterized in that: The removal process of linear structure pixels needs to meet the requirements of high-voltage overhead transmission line devices, and a 5×5 sliding window is set based on morphological opening operation to complete the removal process.

6. The method for automatic separation of power stages without auxiliary data based on laser radar point cloud according to claim 1 is characterized by: After filtering the laser point cloud data, all point clouds within the rectangular frame area are retained.

7. The method for automatic separation of power stages without auxiliary data based on laser radar point cloud according to claim 6 is characterized by: Generate independent storage files in .txt format for all point clouds within the rectangular box area.

Citation Information

Patent Citations

  • Classification method for airborne LIDAR power inspection point cloud

    CN106897686A

  • Airborne point cloud data electric power pole and tower automatic extraction method based on spatial constraint

    CN107680102A

  • Safety assessment method, device, equipment and medium for ultrahigh-pressure maintenance operation personnel

    CN117788902A

  • Laser point cloud-based automatic line tower footing terrain extraction and mapping method

    WO2022062305A1