A method for detecting mountainous power transmission and transformation project slope sliding slag based on remote sensing three-dimensional width change

Abstract

The application discloses a mountainous power transmission and transformation project slope sliding slag detection method based on remote sensing three-dimensional width change, obtains satellite remote sensing images and digital elevation model data of a target area, carries out data preprocessing, generates a slope grid map and processed images, extracts edge lines on the left and right sides of a road from the processed images, generates a vertical profile line along a center line of the road, calculates horizontal plane coordinates of intersection points of each vertical profile line and the edge lines, calculates elevation values of the intersection points according to the digital elevation model data, calculates three-dimensional road widths corresponding to each pair of intersection points according to the horizontal plane coordinates and the elevation values, compares the three-dimensional road widths with a reference width to obtain a width change amount, and carries out slope sliding slag event determination and risk early warning if the width change amount is not less than a width threshold value and a slope value of a corresponding position in the slope grid map is not less than a slope threshold value. The application reduces measurement errors of a traditional planar method in steep slope terrain and improves the accuracy of judgment.

Description

TECHNICAL FIELD

[0001] The present application relates to the technical field of power engineering environment monitoring, in particular to a mountainous power transmission and transformation project slope slurry detection method based on remote sensing three-dimensional width change. BACKGROUND

[0002] In the power transmission and transformation project related to water and soil conservation, the excavation of the construction road in the mountainous area during the construction of the power transmission line is easy to cause the slurry to slide down the slope (i.e. the loose slurry slides down the slope), thereby damaging the surface vegetation and soil structure, causing environmental problems such as soil erosion and water pollution. The traditional manual patrol is low in efficiency and high in risk, and the existing remote sensing monitoring focuses on slurry identification, and does not fully utilize the road shape change characteristics. Secondly, most of the researches use satellite remote sensing image pictures for single discrimination, and do not consider that the position of the slurry sliding down the slope and the piled slurry during the construction must have a height difference. The image pictures only contain two-dimensional coordinate information of the ground, and can only identify the surface disturbance caused by the power transmission and transformation project construction, and cannot accurately identify the slurry sliding down the slope and other problems caused by the power transmission and transformation project construction. This leads to problems such as great difficulty in image recognition and low recognition accuracy.

[0003] The patent application with publication number CN115731471A discloses a construction road quantitative monitoring method based on satellite remote sensing technology, which comprises the following steps: S1: obtaining a remote sensing image Q before road construction and a remote sensing image P after construction through satellite remote sensing technology; S2: taking the area within a certain distance on both sides of the power transmission and transformation line as a buffer zone, and taking the area within 5km outside the power transmission and transformation line as a buffer zone, and taking the buffer zone as a cutting range to cut the remote sensing image Q before construction and the remote sensing image P after construction, to obtain a cut remote sensing image Q1 before construction and a cut remote sensing image P1 after construction; S3: performing difference on the cut remote sensing image Q1 before construction and the cut remote sensing image P1 after construction to obtain an analysis image; S4: identifying the newly added construction road through an RGB space statistical model, a region growing algorithm and manual intervention; and S5: quantitatively calculating the length and width of the newly added construction road through a GIS statistical tool. It can be seen that the above technical solution has the following problems: satellite remote sensing image pictures are used for single discrimination, and the topographic factors are not considered. The recognition accuracy is low, and there is no alternative solution when the local optical image fails, and there is a lack of a road shape-topography-substance characteristic collaborative analysis system. SUMMARY

[0004] The technical problem to be solved by the present application is to provide a mountainous power transmission and transformation project slope slurry detection method based on remote sensing three-dimensional width change, taking the construction road width as the core criterion, combining with the topographic slope analysis, to realize efficient slurry sliding down the slope discrimination.

[0005] In order to solve the above technical problems, the technical scheme adopted by the present application is:

[0006] A mountainous power transmission project slope sliding detection method based on remote sensing three-dimensional width change, comprising the following steps:

[0007] Obtain satellite remote sensing images and digital elevation model data of the target area, and perform data preprocessing on the satellite remote sensing images and digital elevation model data to obtain processed images and generate a slope grid map;

[0008] Extract the edge lines on the left and right sides of the road from the processed images;

[0009] Generate a vertical profile line along the center line of the road, the vertical profile line being a tangent direction straight line segment perpendicular to different points of the center line, calculate the horizontal coordinates of the intersection points of each vertical profile line and the edge lines on the left and right sides of the road, calculate the elevation values of the intersection points according to the digital elevation model data, and calculate the three-dimensional spatial distance of each pair of intersection points to obtain the corresponding three-dimensional road width according to the horizontal coordinates and the elevation values, and compare the three-dimensional road width corresponding to each pair of intersection points with the reference width to obtain the corresponding width change;

[0010] Compare the width change corresponding to each pair of intersection points with the width threshold, and compare the slope value of the corresponding position of the intersection points in the slope grid map with the slope threshold, if the width change corresponding to a pair of intersection points is greater than or equal to the width threshold, and the slope value of the corresponding position of the intersection points in the slope grid map is greater than or equal to the slope threshold, then the specified range of the slope surface above the corresponding position is verified by spectrum to determine whether it is a slope sliding event and to perform risk warning.

[0011] Further, when the satellite remote sensing images and digital elevation model data are preprocessed to obtain processed images and generate a slope grid map, the following steps are included:

[0012] Correct the data of the satellite remote sensing images, superimpose and display the corrected satellite remote sensing images and the digital elevation model data, set a buffer zone along the road center line in the superimposed and displayed images, and then crop the superimposed and displayed images to obtain a cropped digital elevation model data grid image as the processed image;

[0013] Obtain the value of each pixel in the cropped digital elevation model grid image, and calculate the value of the elevation change rate according to the value of each pixel, and then calculate the slope value of each grid point in the digital elevation model grid image according to the value of the elevation change rate.

[0014] Further, when the edge lines on the left and right sides of the road are extracted from the processed images, the following steps are included:

[0015] The processed image is input into a deep learning model to obtain a road probability map. The probability of each pixel in the road probability map is compared with a preset probability threshold. If it is less than the preset probability threshold, the value of the pixel is updated to 0. If it is greater than the preset probability threshold, the value of the pixel is updated to 1. Thus, the road probability map is transformed into a binary raster mask image of non-road and road.

[0016] Morphological operations are performed on the binary raster mask image to obtain an optimized binary raster mask image. Then, the optimized binary raster mask image is converted into a vector polygon. Finally, the single closed vector polygon of the road area is separated into two independent left and right edge lines, and the separated left and right edge lines are smoothed.

[0017] Furthermore, when generating a vertical profile line along the centerline of the road, specifically, center points are selected at specified intervals along the centerline of the road, the corresponding tangent direction is calculated at each center point, and a straight line segment perpendicular to the tangent direction is generated for each tangent as the corresponding vertical profile line.

[0018] Furthermore, when calculating the horizontal coordinates of the intersection points of each vertical profile line with the left and right edge lines of the road, if the intersection points of the vertical profile line with the left or right edge lines of the road are not unique, the intersection point closest to the corresponding center point is selected and its horizontal coordinates are calculated.

[0019] Furthermore, when calculating the elevation value of the intersection point based on the digital elevation model data, the specific steps involve obtaining the vertex coordinates of the grid into which the intersection point falls in the digital elevation model data, and performing elevation interpolation based on the vertex coordinates to obtain the elevation value corresponding to the intersection point. The calculation formula is as follows:

[0020]

[0021] in, This indicates the lower left vertex of the grid into which the intersection point falls. The elevation value, and the lower left vertex The horizontal coordinates are ( , ), This indicates the top-left vertex of the grid into which the intersection point falls. The elevation value, and the top left vertex The horizontal coordinates are ( , ), This indicates the bottom right vertex of the grid into which the intersection point falls. The elevation value, and the lower right vertex The horizontal coordinates are ( , ), a right upper vertex of a grid in which the intersection point falls an elevation value, and a right upper vertex a horizontal plane coordinate of the right upper vertex

[0022] Further, when calculating the three-dimensional spatial distance of each pair of intersection points according to the horizontal plane coordinate and the elevation value, the calculation formula is as follows:

[0023]

[0024] wherein: represents the horizontal plane x-direction coordinate value of the intersection point of the right side edge; represents the horizontal plane y-direction coordinate value of the intersection point of the right side edge; represents the elevation value of the intersection point of the right side edge; represents the horizontal plane x-direction coordinate value of the intersection point of the left side edge; represents the horizontal plane y-direction coordinate value of the intersection point of the left side edge; represents the elevation value of the intersection point of the left side edge.

[0025] Further, when performing spectral verification on the specified range of the slope surface above the corresponding position of the intersection point, if optical images are available, spectral verification is performed on the specified range of the slope surface above the corresponding position of the intersection point, and when optical images are unavailable, SAR image coherent change detection is used to replace spectral verification, after obtaining the spectral verification result or the SAR image coherent change detection result, if the near-infrared band reflectivity 0.3 and the normalized vegetation index NDVI 0.2, it is determined as a down-slope slag sliding event and a high-risk early warning is performed.

[0026] Further, after comparing the width change amount corresponding to each pair of intersection points with the width threshold value and comparing the slope value of the corresponding position of the intersection point in the slope grid map with the slope threshold value, the method further comprises:

[0027] if the width change amount corresponding to a pair of intersection points is less than or equal to the width threshold value but greater than or equal to the first threshold value, and the slope value of the corresponding position of the intersection point in the slope grid map is greater than or equal to the slope threshold value, a medium-risk early warning is performed;

[0028] if the width change amount corresponding to a pair of intersection points is greater than or equal to the second threshold value, the second threshold value is less than the first threshold value, and the corresponding position of the intersection point is in a historical slag sliding point area, a warning is performed.

[0029] ​​The application further provides a mountainous power transmission and transformation project slope sliding slag detection system based on remote sensing three-dimensional width change, which comprises a processor and a computer readable storage medium, the computer readable storage medium stores a computer program, and the computer program is executed by the processor to realize the steps of the mountainous power transmission and transformation project slope sliding slag detection method based on remote sensing three-dimensional width change.

[0030] Compared with the prior art, the application has the advantages that:

[0031] The application calculates the three-dimensional road width by fusing satellite remote sensing images and digital elevation model data, reduces the measurement error of the traditional planar method in steep slope terrain, and solves the error problem of planar measurement in steep slope terrain.

[0032] When the application performs risk early warning, three criteria of width change, slope constraint and spectral verification are combined, the width change greater than the threshold value indicates that an additional accumulation body appears at the roadside, and there may be slag, the slope greater than the threshold value indicates that the slope above is steep and has the condition of sliding slag, and the flat ground expansion false alarm is excluded, and after the two premise conditions are met, the spectral verification is used to determine the slope sliding slag event, so that the accuracy of the construction road slope sliding slag judgment is greatly improved. BRIEF DESCRIPTION OF DRAWINGS

[0033] Figure 1 The method flowchart of the embodiment of the application is shown.

[0034] Figure 2 The image preprocessing flowchart in step S1 of the embodiment of the application is shown.

[0035] Figure 3 The road edge extraction flowchart in step S2 of the embodiment of the application is shown.

[0036] Figure 4 The schematic diagram of calculating three-dimensional spatial distance by using three-dimensional coordinates in step S3 of the embodiment of the application is shown. DETAILED DESCRIPTION

[0037] The application will be further described below in combination with the drawings of the specification and specific preferred embodiments, but the protection scope of the application is not limited thereby.

[0038] The embodiment provides a mountainous power transmission and transformation project slope sliding slag detection method based on remote sensing three-dimensional width change, combines satellite images and digital elevation model (DEM) to calculate the three-dimensional width of the road to eliminate the measurement error caused by the slope terrain, and uses the three-dimensional width change of the road as the core criterion of the slope sliding slag, so as to solve the problem that the traditional slag body recognition method is not accurate in judging the slope sliding slag of the power transmission and transformation construction road and risk control.

[0039] As shown in Figure 1 the method of the embodiment comprises the following steps:

[0040] S1) Collaborative Data Preprocessing: Acquire satellite remote sensing images and digital elevation model data of the target area, perform data preprocessing on the satellite remote sensing images and digital elevation model data to obtain processed images and generate slope raster maps.

[0041] S2) Road edge line extraction: Using a deep learning model, extract the edge lines on the left and right sides of the road from the processed image;

[0042] S3) 3D width calculation: Generate vertical profile lines along the centerline of the road. The vertical profile lines are straight line segments perpendicular to the tangent direction at different points of the centerline. Calculate the horizontal coordinates of the intersection points of each vertical profile line with the edge lines on the left and right sides of the road. Calculate the elevation value of the intersection points based on the digital elevation model data. Calculate the 3D spatial distance of each pair of intersection points based on the horizontal coordinates and elevation values ​​to obtain the corresponding 3D road width. Compare the 3D road width corresponding to each pair of intersection points with the reference width to obtain the corresponding width change.

[0043] S4) Compare the width change corresponding to each pair of intersections with the width threshold, and compare the slope value at the intersection position in the slope raster with the slope threshold. If the width change corresponding to a pair of intersections is greater than or equal to the width threshold, and the slope value at the intersection position in the slope raster is greater than or equal to the slope threshold, then perform spectral verification on the specified range of the slope above the intersection position to determine whether it is a downslope slag dumping event.

[0044] S5) Risk Warning: Based on the judgment results, risk warnings are issued and risk point GIS layers are output.

[0045] The above steps reduce measurement errors in steep slope terrain using traditional planar methods. The proposed triple criterion of "width variation + slope constraint + spectral verification" significantly improves the accuracy of judging slope slippage and debris flow on construction roads in power transmission and transformation projects. This provides stable and reliable technical support for environmental and water conservation monitoring in mountainous power transmission and transformation areas.

[0046] Taking a 220kV transmission line project in Hunan Province as an example, the construction road area is a typical mountainous and hilly region, with a total length of approximately 30km and a maximum slope of 42°. Multiple instances of slag runoff occurred during the rainy season. Each step will be explained in detail using this example.

[0047] In step S1 of this embodiment, when acquiring satellite remote sensing images and digital elevation model (DEM) data of the target area, the higher the resolution of the satellite images and DEM data, the better the recognition effect. To obtain better recognition results, after multiple tests, the satellite images and DEM data should meet the following conditions:

[0048]

[0049] Therefore, the following data for the target area needs to be prepared:

[0050]

[0051] When preprocessing satellite remote sensing imagery and digital elevation model (DEM) data, the first step is to perform one or more of the following on the satellite imagery: atmospheric correction, geometric correction, and image fusion. Next, the satellite imagery and DEM data undergo one or more processing steps, such as spatial benchmark unification, parameter adjustment, and eigenvalue overlay. Then, a 50m buffer zone is generated along the road centerline, and finally, cropping is performed. For example... Figure 2 As shown, it includes the following steps:

[0052] S11) Correcting satellite remote sensing image data, specifically by using software such as ENVI to perform atmospheric correction, geometric correction, and topographic correction on satellite image data in sequence. The purpose is to eliminate the effects of atmospheric scattering and absorption, obtain the true reflectance of ground objects, correct geometric distortion of images, give them accurate geographic coordinates, and eliminate radiation distortion caused by topographic undulations.

[0053] S12) Overlay the corrected satellite remote sensing imagery with digital elevation model data. Specifically, add the corrected satellite imagery and digital elevation model (DEM) data to ArcScene software for overlay display. Then, create an analysis area of ​​a certain width around the point and line features and set a buffer zone along the road centerline. Finally, crop the overlaid imagery according to the study area to obtain the cropped digital elevation model data raster image as the processed image.

[0054] After completing the above steps, generate a slope raster map from the cropped image. The specific steps are as follows:

[0055] S13) Obtain the value of each pixel in the cropped digital elevation model raster image, where the value of each pixel represents the elevation of that point;

[0056] S14) Calculate the elevation change rate (gradient) of the horizontal plane in the X and Y directions based on the value of each pixel. In this embodiment, the third-order inverse distance squared weighted finite difference method is used to traverse each cell of the DEM in a 3x3 window. For a 3x3 window, the elevation values ​​are as follows:

[0057]

[0058] Rate of change in the X direction:

[0059] Rate of change in the Y direction:

[0060] Where 8cellSize is the cell size (resolution) of the DEM raster.

[0061] S15) Calculate the slope value of each grid point in the digital elevation model raster image based on the elevation change rate. In this embodiment, the slope value of each grid point is calculated by the angle between the composite vector of the X and Y direction change rates and the horizontal plane. The calculation formula is as follows:

[0062]

[0063] Where S represents the slope value.

[0064] In step S2 of this embodiment, the deep learning model uses the U-Net model. The U-Net model takes red, green, blue, and near-infrared bands as input and outputs a road probability map. When extracting the edge lines on the left and right sides of the road from the processed image, such as... Figure 3 As shown, it includes the following steps:

[0065] S21) Input the processed image into the U-Net model to obtain a road probability map. Compare the probability of each pixel in the road probability map with a preset probability threshold. If it is less than the preset probability threshold, update the value of the pixel to 0. If it is greater than the preset probability threshold, update the value of the pixel to 1. Thus, the road probability map is transformed into a binary raster mask image of non-road (background, value 0) and road (foreground, value 1).

[0066] S22) Morphological operations are performed on the binary raster mask image to optimize its quality, resulting in an optimized binary raster mask image, which lays the foundation for subsequent vectorization. The morphological operations mainly include closing operations (dilation followed by erosion, used to fill small holes and short breaks inside the mask) and opening operations (erosion followed by dilation, used to eliminate small isolated noise points and smooth boundaries).

[0067] S23) Convert the optimized binarized raster mask image into a vector polygon, including the following steps:

[0068] Edge detection: Using a boundary tracing algorithm (such as the Moore neighborhood tracing algorithm), the binary image is traversed sequentially to accurately find and record the sequence of pixel coordinate points that constitute the outer contour of the road area (and internal holes, if any), thus obtaining pixel-level boundary lines.

[0069] Vector Construction and Simplification: The dense pixel sequence obtained from tracking is constructed into a vector polygon. Subsequently, a vector simplification algorithm (such as the Douglas-Peucker algorithm) is used to simplify the polygon.

[0070] S24) Separate the single closed vector polygon representing the entire road area into two independent edge lines, left and right;

[0071] S25) Smooth the separated left and right edge lines to eliminate the jagged appearance or unnatural jitter that may remain after vector simplification, and obtain a smooth curve that conforms to the shape of the real road.

[0072] Step S3 in this embodiment aims to calculate the three-dimensional road width, including the following steps:

[0073] S31) Generate a vertical profile line along the centerline of the road. Specifically, select a center point at a specified interval (e.g., every 10m) along the centerline of the road, calculate the corresponding tangent direction (azimuth angle α) at ​​each center point, and generate a straight line segment perpendicular to the tangent direction for each tangent as the corresponding vertical profile line. To ensure that the length of the vertical profile line covers the left and right edges, the length of the vertical profile line = road design width * 2.5 (in this embodiment, 6 * 2.5 = 15m). The relevant calculation formula is as follows:

[0074] Let the coordinates of the center point be ( , If the tangent azimuth angle is α (clockwise from true north), then the coordinates of the two endpoints of the vertical profile line are:

[0075] = +L·sin( )

[0076] = +L·cos( )

[0077] = +L·sin( )

[0078] = +L·cos( )

[0079] In the formula: L is the half length of the vertical section line (7.5 meters in this embodiment).

[0080] S32) Calculate the horizontal coordinates of the intersection points of each vertical profile line with the left and right edge lines of the road. Use spatial geometry calculations to find the intersection point P of the vertical profile line and the left edge line. L Horizontal coordinates ( , The intersection point P with the right edge line R Horizontal coordinates ( , If the intersection of the vertical profile line with the left or right edge line of the road is not unique (e.g., the vertical profile line at the bend of the road may intersect with the edge line of the adjacent road), then select the intersection point closest to the corresponding center point and calculate the horizontal coordinates.

[0081] S33) Calculate the elevation value of the intersection point based on the digital elevation model data. Specifically, use bilinear interpolation to obtain the vertex coordinates of the grid in which the intersection point falls in the digital elevation model data, and perform elevation interpolation based on the vertex coordinates to obtain the elevation value corresponding to the intersection point. Let point P ( , The four points falling on the DEM grid ( , ), ( , ), ( , ), ( , If the elevation is between 0 and 1, the formula for calculating the elevation value is as follows:

[0082]

[0083] in, This indicates the lower left vertex of the grid into which the intersection point falls. The elevation value, and the lower left vertex The horizontal coordinates are ( , ), This indicates the top-left vertex of the grid into which the intersection point falls. The elevation value, and the top left vertex The horizontal coordinates are ( , ), This indicates the bottom right vertex of the grid into which the intersection point falls. The elevation value, and the lower right vertex The horizontal coordinates are ( , ), This indicates the top right vertex of the grid into which the intersection point falls. The elevation value, and the upper right vertex The horizontal coordinates are ( , ).

[0084] S34) Calculate the three-dimensional spatial distance of each pair of intersection points based on the horizontal plane coordinates and elevation values, such as... Figure 4 As shown, the calculation formula is as follows:

[0085]

[0086] in: The horizontal x-coordinate value of the intersection point of the right edges; The horizontal y-coordinate value of the intersection point of the right edges; The elevation value of the intersection point representing the right edge; The horizontal x-coordinate value of the intersection point of the left edges; The horizontal y-coordinate value of the intersection point of the left edges; This represents the elevation value of the intersection point on the left edge.

[0087] S35) Width change analysis compares the calculated three-dimensional road width with the baseline width (design value or historical value) to obtain the width change amount:

[0088]

[0089] in, This dataset represents the baseline width of the road at each center point. In this embodiment, the baseline width is based on the data from the initial construction phase. This dataset represents the 3D road width corresponding to each center point.

[0090] For the example construction road, some calculation results are shown in the table below. The station number in the table represents the station number corresponding to a pair of intersections, and the slope in the table is the slope value of the grid point at the corresponding position of the pair of intersections or the center point of the pair of intersections in the slope grid map.

[0091]

[0092] In step S4 of this embodiment, after multiple tests, the range of values ​​for the width threshold and the slope threshold must meet the following conditions:

[0093]

[0094] in, Indicates the width threshold. This indicates the slope threshold.

[0095] In step S4, when performing spectral verification on a specified range of the slope above the intersection point, the intersection point or its corresponding center point is used as the determination point, and spectral verification is performed on a range of 50m above the determination point. This includes the following steps:

[0096] S41) When optical images are available, perform spectral verification on a slope within 50m above the determination point;

[0097] S42) When optical images are unavailable, SAR image coherence change detection is used as an alternative to spectral verification.

[0098] After obtaining the spectral verification results or the SAR image coherence change detection results, the slope sludge chuting event must simultaneously meet the following conditions: near-infrared reflectance. 0.3 and Normalized Difference Vegetation Index (NDVI) 0.2. The purpose is to eliminate interference from vegetation factors and to identify dry slag.

[0099] The core reason for setting near-infrared reflectance > 0.3 and NDVI < 0.2 as the criteria for judging construction waste in this embodiment is that the spectral characteristics of construction waste are fundamentally different from those of natural ground surfaces. The specific scientific mechanism is as follows:

[0100] The physical mechanisms underlying near-infrared reflectance > 0.3 are shown in the table below:

[0101]

[0102] In this embodiment, it was verified that the near-infrared reflectance of sandstone slag is 0.35-0.45 (dry state), which is significantly higher than that of surrounding vegetation (0.1-0.25) and moist soil (0.15-0.25). A threshold of 0.3 can effectively distinguish slag from natural ground surface.

[0103] The biological mechanisms of NDVI < 0.2 are shown in the table below:

[0104]

[0105] In this embodiment, it was verified that the chlorophyll absorption rate of healthy vegetation in the red light band is >90%, while the near-infrared reflectance can reach 40-60%, resulting in an NDVI value often greater than 0.6. In slag heap areas, due to the lack of vegetation cover, the calculated value is typically negative to between 0.2.

[0106] In step S5 of this embodiment, risk classification is performed on the relevant information after the slope and slag discharge determination, including:

[0107] If the width change corresponding to a pair of intersection points is greater than or equal to a width threshold, and the slope value at the corresponding position of the intersection point in the slope raster is greater than or equal to a slope threshold, and the spectral verification result of the specified range of the slope above the corresponding position of the intersection point satisfies: near-infrared band reflectance 0.3 and Normalized Difference Vegetation Index (NDVI) If the value is 0.2, it is determined to be a slope runoff event and a high-risk warning is issued.

[0108] If the width change corresponding to a pair of intersection points is less than or equal to the width threshold but greater than or equal to the first threshold, and the slope value at the location corresponding to the intersection point in the slope raster is greater than or equal to the slope threshold, then a medium-risk warning is issued.

[0109] If the width change corresponding to a pair of intersection points is greater than or equal to the second threshold, the second threshold is less than the first threshold, and the location corresponding to the intersection point is in the historical slag dumping area, then a warning is issued.

[0110] In this embodiment, the width threshold The maximum value of 2m is taken within its range, which is the slope threshold. Taking the minimum value of 25° within its range, with a first threshold of 1.5m and a second threshold of 1m, the risk classification standard is as follows:

[0111]

[0112] After determining the risk level, an early warning message is generated and pushed to relevant units. Corresponding measures are then taken, as follows:

[0113]

[0114] For the power transmission line project in the example, the detection results obtained by performing the above steps are as follows:

[0115]

[0116] As can be seen, through the above steps, the accuracy rate of slope runoff and slag discharge confirmation reaches 88.2%, which is higher than the accuracy and efficiency of traditional methods, providing a stable and reliable technical guarantee for the field of power transmission and transformation environmental and water conservation monitoring in mountainous areas.

[0117] Furthermore, this embodiment also proposes a slope sludge chute detection system for mountain power transmission and transformation projects based on remote sensing three-dimensional width variation, including a processor and a computer-readable storage medium. The computer-readable storage medium stores a computer program, which is executed by the processor to implement the steps of the slope sludge chute detection method for mountain power transmission and transformation projects based on remote sensing three-dimensional width variation described in this embodiment.

[0118] In summary, this invention provides a method for detecting slag chutes on slopes of power transmission and transformation projects in mountainous areas based on three-dimensional width changes from remote sensing. The core innovation lies in the three-dimensional width calculation model and the multi-criteria system. Compared with traditional methods, it has the following advantages: (1) The three-dimensional width calculation reduces the measurement error of traditional planar methods on steep slopes by integrating DEM elevation data, thus solving the error problem of planar measurement on steep slopes. (2) The proposed triple criteria of "width change + slope constraint + spectral verification" greatly improves the accuracy of judging slag chutes on construction roads, such as distinguishing between slag burial and temporary stockpiling. (3) It supports multi-cloud and multi-source data collaborative processing, ensuring monitoring continuity and providing stable and reliable technical support for the field of environmental and water conservation monitoring of power transmission and transformation projects in mountainous areas.

[0119] Those skilled in the art will understand that embodiments of this application can be provided as methods, systems, or computer program products. Therefore, this application can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, this application can take the form of a computer program product embodied on one or more computer-readable storage media (including, but not limited to, disk storage, CD-ROM, optical storage, etc.) containing computer-usable program code. This application is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of this application. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, create a machine for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The computer program instructions may also be stored in a computer-readable storage medium that can direct a computer or other programmable data processing device to operate in a particular manner, such that the instructions stored in the computer-readable storage medium produce an article of manufacture including instruction means, which are implemented in a process Figure 1 One or more processes and / or boxes Figure 1 The functions specified in one or more boxes. These computer program instructions may also be loaded onto a computer or other programmable data processing apparatus to cause a series of operational steps to be performed on the computer or other programmable apparatus to produce a computer-implemented process, thereby providing instructions that execute on the computer or other programmable apparatus for implementing the process. Figure 1 One or more processes and / or boxes Figure 1 The steps of the function specified in one or more boxes.

[0120] The above description is merely a preferred embodiment of the present invention. The scope of protection of the present invention is not limited to the above embodiments. All technical solutions falling within the scope of the present invention's concept are within the scope of protection of the present invention. It should be noted that for those skilled in the art, any improvements and modifications made without departing from the principles of the present invention should also be considered within the scope of protection of the present invention.

Claims

1. A method for detecting slope sliding of slag in mountainous power transmission and transformation projects based on remote sensing three-dimensional width change, characterized in that, The method comprises the following steps: Satellite remote sensing images and digital elevation model data of a target area are acquired, and the satellite remote sensing images and the digital elevation model data are preprocessed to obtain processed images and generate a slope grid map; Edge lines on the left and right sides of a road are extracted from the processed images; Perpendicular profile lines are generated along the center line of the road, the perpendicular profile lines are straight line segments perpendicular to the tangent directions of different points on the center line, the horizontal coordinates of the intersection points of each perpendicular profile line and the edge lines on the left and right sides of the road are calculated, the elevation values of the intersection points are calculated according to the digital elevation model data, the three-dimensional spatial distances of each pair of intersection points are calculated according to the horizontal coordinates and the elevation values to obtain corresponding three-dimensional road widths, and the corresponding three-dimensional road widths of each pair of intersection points are compared with a reference width to obtain corresponding width variation amounts; The width variation amounts of each pair of intersection points are compared with a width threshold value, and the slope values of the positions corresponding to the intersection points in the slope grid map are compared with a slope threshold value, if the width variation amount of a pair of intersection points is greater than or equal to the width threshold value, and the slope value of the position corresponding to the intersection points in the slope grid map is greater than or equal to the slope threshold value, then a specified range of the slope surface above the position corresponding to the intersection points is subjected to spectral verification to determine whether it is a down-slope slurry sliding event and perform a risk warning.

2. The method according to claim 1, wherein the method is characterized by, When the satellite remote sensing images and the digital elevation model data are preprocessed to obtain processed images and generate a slope grid map, the following steps are included: The data of the satellite remote sensing images are corrected, the corrected satellite remote sensing images are combined with the digital elevation model data for display, a buffer zone is set along the center line of the road in the combined display image, and then the combined display image is cropped to obtain a cropped digital elevation model data grid image as the processed image; The values of each pixel in the cropped digital elevation model grid image are acquired, and the values of the elevation change rates are calculated according to the values of each pixel, and then the slope values of each grid point in the digital elevation model grid image are calculated according to the values of the elevation change rates.

3. The method according to claim 2, wherein the method is characterized by, When the edge lines on the left and right sides of the road are extracted from the processed images, the following steps are included: The processed image is input into a deep learning model to obtain a road probability map, the probability of each pixel point in the road probability map is compared with a preset probability threshold value, if the probability is less than the preset probability threshold value, the value of the pixel point is updated to 0, if the probability is greater than the preset probability threshold value, the value of the pixel point is updated to 1, so as to convert the road probability map into a binary grid mask image of non-road and road; Morphological operations are performed on the binary grid mask image to obtain an optimized binary grid mask image, and then the optimized binary grid mask image is converted into a vector polygon, finally, a single closed vector polygon of the road region is separated into left and right two independent edge lines, and the separated left and right edge lines are subjected to smoothing processing.

4. The method according to claim 1, wherein the method is characterized by, When the perpendicular profile lines are generated along the center line of the road, the center points are selected at a specified interval along the center line of the road, the corresponding tangent directions are calculated at each center point, and a straight line segment perpendicular to the tangent direction is generated as the corresponding perpendicular profile line for each tangent line.

5. The method according to claim 4, wherein the method is characterized by, When calculating the horizontal plane coordinates of the intersection points of each vertical profile line and the edge lines on the left and right sides of the road, if the intersection points of the vertical profile line and the edge line on the left or right side of the road are not unique, the intersection point closest to the corresponding center point is selected and the horizontal plane coordinates are calculated.

6. The method according to claim 1, wherein the method is characterized by, When calculating the elevation values of the intersection points according to the digital elevation model data, specifically, the vertex coordinates of the grid in which the intersection points fall are obtained, and elevation interpolation is performed according to the vertex coordinates to obtain the elevation values corresponding to the intersection points, and the calculation formula is as follows: wherein represents the elevation value of the lower left corner of the grid into which the intersection falls, and the horizontal plane coordinates of the lower left corner are (x, y), represents the elevation value of the upper left corner of the grid into which the intersection falls, and the horizontal plane coordinates of the upper left corner are (x, y + 1), represents the elevation value of the lower right corner of the grid into which the intersection falls, and the horizontal plane coordinates of the lower right corner are (x + 1, y), represents the elevation value of the upper right corner of the grid into which the intersection falls, and the horizontal plane coordinates of the upper right corner are (x + 1, y + 1), .​​​​​​​​​​​ 7. The method according to claim 1, wherein the method is characterized by, When calculating the three-dimensional space distance of each pair of intersection points according to the horizontal plane coordinates and the elevation values, the calculation formula is as follows: wherein: a horizontal plane x-direction coordinate value of the intersection point representing the right side edge; a horizontal plane y-direction coordinate value of the intersection point representing the right side edge; an elevation value of the intersection point representing the right side edge; a horizontal plane x-direction coordinate value of the intersection point representing the left side edge; a horizontal plane y-direction coordinate value of the intersection point representing the left side edge; an elevation value of the intersection point representing the left side edge.

8. The method according to claim 1, wherein the method is characterized by, If the slope above the intersection corresponding position is designated, the spectrum verification is performed on the designated range of the slope above the intersection corresponding position, and the spectrum verification is performed on the designated range of the slope above the intersection corresponding position when the optical image is available, and the SAR image coherent change detection is used to replace the spectrum verification when the optical image is unavailable, and after obtaining the spectrum verification result or the SAR image coherent change detection result, if the near-infrared band reflectivity 0.3 and the normalized vegetation index NDVI 0.2, it is determined as the down-slope slag sliding event and high-risk early warning is performed.

9. The method according to claim 1, wherein the method is characterized by, After comparing the width change amount corresponding to each pair of intersection points with the width threshold value and comparing the slope value of the intersection point corresponding position in the slope grid map with the slope threshold value, the method further comprises: If the width change amount corresponding to a pair of intersection points is less than or equal to the width threshold value but greater than or equal to the first threshold value, and the slope value of the intersection point corresponding position in the slope grid map is greater than or equal to the slope threshold value, a medium-risk early warning is performed; If the width change amount corresponding to a pair of intersection points is greater than or equal to the second threshold value, the second threshold value is less than the first threshold value, and the intersection point corresponding position is in the historical slag sliding point area, an alert is performed.

10. A mountainous power transmission and transformation project slope sliding slag detection system based on remote sensing three-dimensional width change, characterized in that, The method comprises a processor and a computer readable storage medium, the computer readable storage medium stores a computer program, and the computer program is executed by the processor to realize the steps of the mountainous area power transmission and transformation project slope sliding slag detection method based on remote sensing three-dimensional width change according to any one of claims 1-9.

Images