Satellite remote sensing-based high-risk assessment methods, devices, and media for power transmission line trees

Abstract

This invention relates to the field of remote sensing measurement technology, specifically to a method, device, and medium for high-risk assessment of trees along power transmission lines based on satellite remote sensing. It acquires red and near-infrared bands from remote sensing images to identify vegetation and outputs the spatial distribution results of its intersections with power transmission lines. Based on these results and line data, safe distances are calibrated and corrected using terrain elevation. Trees are identified, and their heights and clearance distances to the lines are calculated. The clearance distances are compared with safety standards to classify high, medium, and low risk levels, generating a high-risk assessment result for trees along power transmission lines. This invention extracts spectral features from remote sensing images to characterize vegetation distribution, constructs dynamic safe zones by combining the location of power transmission lines with vegetation distances, and integrates terrain and tree canopy elevation data to establish a model. It quantifies clearance parameters to generate risk levels, improving adaptability to complex terrain and the accuracy of automatic risk point identification, enhancing early warning reliability, and promoting intelligent and precise decision-making in power transmission line inspections.

Description

Technical Field

[0001] This invention relates to the field of remote sensing measurement technology, and in particular to a method, apparatus and medium for high-risk assessment of power transmission line trees based on satellite remote sensing. Background Technology

[0002] The field of remote sensing measurement technology includes the measurement and monitoring of ground targets using equipment such as satellites, drones, or aerial platforms. The core technology of this field is to remotely acquire images and data of the ground or objects to assess and analyze terrain, environment, resources, and meteorology. With its non-contact advantage, remote sensing technology can cover a wide area and acquire high-precision data in real time. It is widely used in agriculture, environmental monitoring, urban planning, resource management, and many other fields. The key to remote sensing measurement technology lies in the combination of image data processing, spatial analysis, and geographic information systems, so that the acquired remote sensing data can effectively reflect various changes in the target area.

[0003] Among them, the satellite remote sensing-based method for assessing the high risk of trees along power transmission lines involves acquiring height information of trees around power transmission lines using remote sensing technology and combining this information with geographic information systems for data analysis. This method assesses the potential risks that trees may pose to power transmission lines, primarily addressing the limitations of traditional tree height measurement methods. Traditional methods rely on manual on-site measurements, which are not only inefficient and pose safety hazards, but also have limited coverage areas. Therefore, this method uses satellite remote sensing imagery data to acquire tree height variation information over a wide area. By analyzing tree growth status using remote sensing and topographic data, it provides a quantitative assessment of the high risk of trees along power transmission lines. This approach can effectively assess the potential threat of trees to power transmission lines over a large area, improving the scientific rigor and accuracy of power transmission line maintenance.

[0004] Existing technologies rely on single remote sensing data to measure the height of trees around transmission lines, lacking a multi-dimensional spatial data fusion mechanism. This leads to decreased identification accuracy in complex terrain and mixed vegetation areas, making it difficult to fully reflect the spatial relationship between vegetation distribution and transmission lines. Furthermore, dynamic safety zone models have not been established, and risk assessment relies on static distance thresholds, resulting in distorted results and unstable risk level classification. The lack of correction for terrain undulations and vegetation density changes increases the error in clearance distance calculation, causing risk identification to lag behind actual changes. In large-scale power transmission scenarios, these technologies suffer from slow monitoring response, insufficient data resolution, and low accuracy in safety assessment. Summary of the Invention

[0005] To address the technical problems existing in the prior art, embodiments of the present invention provide a method, apparatus, and medium for high-risk assessment of power transmission line trees based on satellite remote sensing. The technical solution is as follows:

[0006] A high-risk assessment method for power transmission line trees based on satellite remote sensing includes the following steps:

[0007] The system acquires red and near-infrared bands from satellite remote sensing images, extracts vegetation spectral reflectance values, identifies vegetation pixels based on the difference in reflectance between red and near-infrared bands, classifies vegetation types using vegetation indices, extracts vegetation areas that intersect with power transmission lines using spatial coordinates, and generates remote sensing vegetation spatial distribution results.

[0008] Based on the vegetation area information in the remote sensing vegetation distribution results, spatial pairing is performed in conjunction with the transmission line route and location to determine the position and distance between the conductor and the vegetation, identify risk areas, and dynamically adjust the range according to the characteristics of the neighboring vegetation to generate safe distance areas.

[0009] Based on the spatial location of vegetation within the safe distance area, obtain the topographic elevation data corresponding to the vegetation area, analyze the elevation changes of the vegetation area and the surrounding terrain, compare the elevation differences of adjacent pixels, identify areas exceeding the threshold and correct them, and generate a terrain correction dataset.

[0010] Based on the spatial location and elevation data of vegetation areas in the terrain correction dataset, the tree canopy apex of the vegetation area is identified, and the tree height is calculated by comparing it with the ground elevation difference. The clearance distance between the tree canopy apex and the power transmission line is obtained, trees or areas with insufficient clearance are identified, and the clearance analysis results between trees and power transmission lines are generated.

[0011] As a further aspect of the present invention, the remote sensing vegetation spatial distribution results include vegetation spatial location, vegetation type classification, and information on the crossing area of ​​power transmission lines; the safe distance area includes the spatial orientation of power transmission lines, relative distance of vegetation, and risk zone boundary; the terrain correction dataset includes vegetation area elevation values, terrain undulation parameters, and elevation correction coefficients; and the tree and power transmission line clearance analysis results include tree height, crown apex elevation, and clearance distance data.

[0012] As a further aspect of the present invention, the step of obtaining the spatial distribution results of remotely sensed vegetation is as follows:

[0013] The system acquires red and near-infrared reflectance data from satellite remote sensing images. It extracts red and near-infrared reflectance values ​​for each pixel in the image, calculates the difference between the two reflectance values ​​for the same pixel, classifies the pixels according to the set red and near-infrared difference threshold, selects a set of pixels that meet the vegetation characteristics, and generates a vegetation pixel reflectance difference set.

[0014] Based on the vegetation pixel reflectance difference set, the red light to near-infrared reflectance ratio value is calculated for each pixel. The type is divided according to the ratio value and the set vegetation reflectance threshold range. The pixel distribution area of ​​different vegetation types is identified respectively. The coordinate positions of each type of pixel are spatially registered to generate a vegetation type distribution coefficient set.

[0015] The set of vegetation type distribution coefficients is called, and the spatial coordinates of the pixels are spatially correlated based on the spatial coordinate data of the transmission lines. The intersection and proximity relationships are detected, and the intersection area is extracted based on the spatial distance threshold between the pixel and the transmission line. The relevant coordinates and vegetation type information are integrated to generate the spatial distribution results of remote sensing vegetation.

[0016] As a further aspect of the present invention, the step of obtaining the safe distance area is as follows:

[0017] Extract vegetation area information from the remote sensing vegetation spatial distribution results, obtain data on power transmission lines, including line direction and location, match these data with the geographic coordinates of the vegetation area, calculate the coordinate offset value between the power transmission line and the vegetation area, correct the line direction, and generate a set of line spatial offset parameters.

[0018] Based on the set of line spatial offset parameters, the relative position of the transmission line and the vegetation area is calculated, the shortest distance between the line and each vegetation area is measured, and these distances are compared with the set safety distance. Areas below the safety threshold are extracted to obtain the set of line vegetation distance coefficients.

[0019] Based on the set of vegetation distance coefficients for the transmission line, and combined with the spatial location, vegetation type and vegetation density of the adjacent vegetation areas, the density rate and spatial overlap ratio of each neighborhood are calculated. The values ​​are compared with the distance coefficients of the transmission line to dynamically adjust the range of potential risk areas and generate safe distance areas.

[0020] As a further aspect of the present invention, the step of obtaining the terrain correction dataset is as follows:

[0021] Based on the spatial location of the vegetation area within the safe distance, the topographic elevation data of the corresponding area is obtained, the elevation value of each pixel in the topographic data is extracted, the elevation data is matched with the spatial coordinates of the vegetation area, the elevation difference between the pixels in the vegetation area and the surrounding terrain is calculated, and a set of topographic elevation difference values ​​is generated.

[0022] Based on the terrain elevation difference value set, the elevation differences of adjacent vegetation area pixels are compared, areas with elevation differences exceeding a set threshold are marked, the elevation values ​​of the marked areas are corrected, and an elevation correction dataset is generated.

[0023] Based on the elevation correction dataset, the elevation changes of the corrected area and the surrounding terrain are compared to identify the accuracy of the correction results. The corrected elevation data is then combined with vegetation area information to generate a terrain correction dataset.

[0024] As a further aspect of the present invention, the steps for obtaining the clearance analysis results between trees and transmission lines are as follows:

[0025] Based on the spatial location and elevation data of vegetation areas in the terrain correction dataset, the changing trend of elevation values ​​of each vegetation cell is detected, the range of elevation differences of cells in the same area is determined, the set of cells with continuously rising elevations is identified and the center position of the set is determined, the center position is marked as the position of the tree crown apex, and a set of tree crown apex elevation values ​​is generated.

[0026] The tree canopy apex elevation value set and the lowest ground elevation value in the same area are called, the numerical difference between the two is calculated and recorded as the height value of each tree, a height sequence is established based on the tree spatial coordinates, vegetation units with height differences exceeding a set threshold are filtered out, and a tree height difference sequence is generated.

[0027] Based on the tree height difference sequence and the spatial coordinates of the transmission line, the vertical distance between the top of each tree crown and the lowest conductor point of the corresponding transmission line is calculated. The distance value is compared with the safety clearance standard to determine the tree number and the area where the distance is lower than the safety requirement, and the tree and transmission line clearance analysis results are generated.

[0028] As a further aspect of the present invention, the method further includes:

[0029] Based on the clearance distance data in the tree and transmission line clearance analysis results, the clearance distance of each vegetation area is compared with the set safety standard. The risk level is divided according to the results: insufficient clearance is high risk, slightly below the standard is medium risk, and meeting the standard is low risk. The risk level and related data are combined to generate the high risk assessment result of the transmission line tree.

[0030] The high-risk assessment results of the transmission line trees include risk level ranges, risk area distribution, and risk assessment indicators.

[0031] As a further aspect of the present invention, the steps for obtaining the high-risk assessment results of the transmission line tree are as follows:

[0032] Based on the clearance distance data in the clearance analysis results of trees and transmission lines, the clearance distance values ​​of all pixels in each vegetation area are detected. The clearance distance of each pixel is compared item by item according to the set safety standard threshold. The difference between the clearance distance and the safety standard threshold is calculated. The clearance distance status in the area is judged based on the difference, and a clearance distance difference set is generated.

[0033] The numerical data in the set of airspace distance differences are called up. Based on the minimum threshold of the safety standard and the range of differences, areas with airspace distance less than or equal to the minimum threshold of the safety standard are selected and set as high-risk areas. Areas with airspace distance differences between the safety standard and the minimum threshold are classified as medium-risk areas, and risk level division interval value groups are generated.

[0034] Based on the remaining data in the risk level division interval value group and the clearance distance difference set, the areas with clearance distance greater than or equal to the safety standard threshold are determined. The risk level values ​​of each area are merged with the corresponding spatial coordinates and area numbers, and the risk value distribution is integrated according to the level interval to generate the high risk assessment result of the transmission line tree.

[0035] As a further aspect of the present invention, after obtaining the terrain correction dataset, the elevation differences of different vegetation areas are classified and identified. Areas with similar elevation change trends, differences, and terrain characteristics are grouped into the same category and assigned corresponding elevation pattern labels. In the subsequent risk assessment process, the safety distance and risk range are adjusted based on the elevation pattern labels.

[0036] As a further aspect of the present invention, after obtaining the results of the clearance analysis of the trees and transmission lines, the clearance distances of different vegetation areas are classified and identified. Areas with similar clearance distances, vegetation types and densities are grouped into the same category and assigned corresponding risk level labels. In subsequent risk assessments, differentiated risk control measures are implemented based on the risk level labels.

[0037] A high-risk assessment device for power transmission line trees based on satellite remote sensing, comprising:

[0038] The first determination module is used to acquire the red and near-infrared bands of satellite remote sensing images, extract vegetation spectral reflectance values, identify vegetation pixels based on the difference in reflectance between the red and near-infrared bands, classify and determine vegetation types using vegetation indices, extract vegetation areas that intersect with power transmission lines by combining spatial coordinates, and generate remote sensing vegetation spatial distribution results.

[0039] The second determining module is used to determine the position and distance between the conductor and the vegetation based on the vegetation area information in the remote sensing vegetation distribution results, combined with the direction and location of the transmission line, to identify risk areas, dynamically adjust the range according to the characteristics of the neighboring vegetation, and generate a safe distance area.

[0040] The correction module is used to obtain the topographic elevation data corresponding to the vegetation area based on the spatial location of the vegetation within the safe distance area, analyze the elevation changes of the vegetation area and the surrounding terrain, compare the elevation difference of adjacent pixels, identify areas that exceed the threshold and correct them, and generate a terrain correction dataset.

[0041] The analysis module is used to identify the tree crown apex in the vegetation area based on the spatial location and elevation data of the vegetation area in the terrain correction dataset, calculate the tree height by comparing it with the ground elevation difference, obtain the clearance distance between the tree crown apex and the transmission line, identify trees or areas with insufficient clearance, and generate the tree and transmission line clearance analysis results.

[0042] A computer-readable storage medium storing a computer program thereon, which, when executed by a processor, implements the steps of any of the methods described above. The beneficial effects of the technical solutions provided by the embodiments of the present invention include at least the following:

[0043] 1. In this invention, by extracting and differentiating the spectral features of multi-source remote sensing images by region, a refined spatial characterization of vegetation distribution can be achieved, effectively improving the accuracy and detail richness of vegetation distribution data. By combining the spatial positioning of transmission lines with the distance relationship between vegetation, a dynamic safety zone can be constructed, thereby reflecting the spatial interaction between transmission lines and surrounding vegetation in real time and improving the response sensitivity of safety monitoring.

[0044] 2. In this invention, by superimposing terrain elevation change correction and tree canopy elevation extraction, a multi-dimensional data fusion model is formed, which can accurately describe the relationship between vegetation and terrain. Combining terrain factors and tree characteristics, it not only strengthens the spatial matching degree between vegetation areas and power transmission lines, but also greatly enhances the monitoring adaptability in complex terrain environments and improves the comprehensiveness and accuracy of risk assessment.

[0045] 3. In this invention, a continuous risk level distribution is generated by quantitative calculation based on the clearance space parameters, realizing the dynamic identification and risk prediction of the spatial relationship between vegetation and lines in the transmission channel. This data modeling method optimizes risk warning, reduces the accumulation of errors in monitoring data, enhances the accuracy of automatic identification of potential risk points, and improves the intelligence level of transmission line inspection and the accuracy of data decision-making. Attached Figure Description

[0046] Figure 1 This is a flowchart of the method of the present invention;

[0047] Figure 2 This is a flowchart illustrating the process of obtaining the spatial distribution results of remotely sensed vegetation according to the present invention.

[0048] Figure 3 This is a flowchart illustrating the process of obtaining the safe distance area according to the present invention.

[0049] Figure 4 This is a flowchart illustrating the process of obtaining the terrain correction dataset for this invention.

[0050] Figure 5 This is a flowchart illustrating the process of obtaining the clearance analysis results between trees and power transmission lines in this invention.

[0051] Figure 6 This is a flowchart illustrating the process of obtaining the high-risk assessment results of the power transmission line tree according to the present invention. Detailed Implementation

[0052] The technical solution of the present invention will now be described with reference to the accompanying drawings.

[0053] In embodiments of the present invention, words such as "exemplarily," "for example," etc., are used to indicate that something is an example, illustration, or description. Any embodiment or design described as "exemplary" in the present invention should not be construed as being more preferred or advantageous than other embodiments or designs. Specifically, the use of the word "exemplary" is intended to present the concept in a concrete manner. Furthermore, in embodiments of the present invention, the meaning expressed by "and / or" can be both, or either one.

[0054] In the embodiments of this invention, the terms "image" and "picture" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning. Similarly, the terms "of," "corresponding (relevant)," and "corresponding" may sometimes be used interchangeably. It should be noted that, without emphasizing the distinction between them, they convey the same meaning.

[0055] In this embodiment of the invention, sometimes a subscript such as W1 may be written in a non-subscript form such as W1. When the difference is not emphasized, the meaning they express is the same.

[0056] To make the technical problems, technical solutions and advantages of the present invention clearer, a detailed description will be given below in conjunction with the accompanying drawings and specific embodiments.

[0057] Please see Figure 1 This invention provides a technical solution: a high-risk assessment method for power transmission line trees based on satellite remote sensing, comprising the following steps:

[0058] The system acquires red and near-infrared bands from satellite remote sensing images, extracts spectral reflectance values ​​that reflect vegetation characteristics, analyzes the spectral reflectance value of each pixel in the image, and combines the reflectance characteristics of vegetation in each spectral band to determine whether a pixel belongs to a vegetation type. Based on the difference in spectral reflectance values ​​between the red and near-infrared bands, vegetation areas are identified, and pixels that meet the vegetation characteristics are marked as vegetation areas. The vegetation index is used to classify each pixel in the image and determine the vegetation type. Based on the spatial coordinates, the system outputs the vegetation areas that intersect with power transmission lines, generating the spatial distribution results of remote sensing vegetation.

[0059] Based on the vegetation area information extracted from the spatial distribution results of remote sensing vegetation, the transmission line route and location information in the transmission line data are obtained. The transmission lines and vegetation areas are spatially paired to determine and mark the relative position and distance relationship between the vegetation area and the transmission line conductor. Potential risk areas are identified based on the distance relationship. By comparing the spatial position, vegetation type and vegetation density of neighboring vegetation areas, the risk interval is dynamically adjusted to generate a safe distance area.

[0060] Based on the spatial location of vegetation areas within the safe distance, the topographic elevation data corresponding to the vegetation areas is obtained, the elevation changes of the vegetation areas and the surrounding terrain are analyzed, the elevation data of adjacent vegetation area pixels are compared, areas with elevation differences exceeding the set threshold are identified, and elevation difference correction is performed to generate a terrain correction dataset.

[0061] Based on the spatial location and elevation data of vegetation areas from the terrain correction dataset, trees in the vegetation areas are identified and the position of the tree crown apex is located. The height of the tree is obtained by the difference between the elevation of the tree crown apex and the elevation of the lowest point on the ground. The clearance distance between the tree crown apex and the transmission line is calculated. Trees or areas with clearance distances lower than the set safety requirements are identified and the clearance status of the corresponding trees or areas is recorded. The tree and transmission line clearance analysis results are generated.

[0062] Based on the clearance distance data from the tree and transmission line clearance analysis, the clearance distance of each vegetated area is compared with the set safety standards. According to the comparison results, the vegetated areas are divided into different risk level intervals. When the clearance distance is less than or equal to the minimum requirement of the safety standard, it is classified as a high-risk area. When the clearance distance is lower than the minimum requirement of the safety standard, but there is still some margin in the distance standard, it is classified as a medium-risk area. When the clearance distance is greater than or equal to the requirement of the safety standard, it is classified as a low-risk area. The risk level of each area is combined with the corresponding risk data to generate the high-risk assessment result of the transmission line tree.

[0063] The results of remote sensing vegetation spatial distribution include vegetation spatial location, vegetation type classification, and information on areas where transmission lines cross. The safe distance area includes the spatial orientation of transmission lines, relative distances to vegetation, and risk zone boundaries. The terrain correction dataset includes elevation values ​​of vegetation areas, terrain undulation parameters, and elevation correction coefficients. The results of tree and transmission line clearance analysis include tree height, crown elevation, and clearance data. The results of transmission line tree height risk assessment include risk level intervals, risk area distribution, and risk assessment indicators.

[0064] Please see Figure 2 The steps for obtaining the spatial distribution results of remotely sensed vegetation are as follows:

[0065] The spectral reflectance values ​​in satellite remote sensing image data are obtained. The pixel reflectance data of red light, near-infrared and blue light bands are extracted in sequence according to the image spectral bands. The multi-band reflectance values ​​under the same geographic coordinates are numerically normalized. The reflectance difference between each band is calculated. The vegetation attribute of the pixel is determined based on the numerical change of the reflectance difference between the red light and near-infrared bands, and a set of band reflectance difference values ​​is generated.

[0066] To acquire spectral reflectance values ​​from satellite remote sensing imagery, a Landsat 9C2L2 level image product acquired on a specific date is used. This image has a spatial resolution of 30 meters, and the pixel values ​​are corrected surface reflectance. The red band (fourth band) and near-infrared band (fifth band) are selected as data sources. For any pixel within the transmission line corridor, the red band reflectance value is extracted. and near-infrared band reflectivity value ,For example:

[0067] Pixel P1 (116.31°, 39.92°) It is 0.08. The value is 0.52 for pixel P2 (116.32°, 39.93°). It is 0.25. The value is 0.30, which means that the values ​​under the same geographic coordinates are... and Pair them up and calculate the direct difference in reflectivity between the two bands. Taking pixel P1 as an example, its difference The difference of pixel P2 The magnitude of the difference reflects the probability and health status of vegetation cover. Pixels with a difference greater than 0.4 are initially judged as high-probability vegetation areas (e.g., P1), and pixels with a difference close to 0 are initially judged as non-vegetated areas (e.g., P2). This process is repeated for all pixels in the transmission line corridor image. and Extraction and difference calculation of the geographic coordinates of all pixels, , The calculated reflectance differences are integrated to generate a band reflectance difference set. Before applying satellite remote sensing image data, strict preprocessing is required, including format parsing, radiometric correction, geometric correction, orthorectification, coordinate projection transformation, image fusion, mosaicking, and cropping. This ensures that the image data meets quality requirements such as panchromatic space resolution less than 1m, multispectral resolution less than 4m, cloud cover less than 20%, and side sway angle less than ±15 degrees. The CGCS2000 coordinate system is also uniformly adopted to achieve accurate matching and coordinate correspondence between remote sensing images and the location of power transmission and transformation engineering lines.

[0068] Based on the band reflectance difference set, the normalized vegetation index value is calculated for each pixel. The red band reflectance value and the near-infrared band reflectance value are used as inputs to perform numerical comparison. The obtained vegetation index value is compared with the set vegetation index threshold. Pixels with index values ​​higher than the threshold are selected and marked as vegetation areas to obtain the vegetation index determination result.

[0069] Based on the band reflectance difference set, the red band reflectance value and near-infrared band reflectance value of each pixel are retrieved to calculate the Normalized Difference Vegetation Index (NDVI) value for each pixel. The calculation formula is as follows: Taking pixel P1 in S101 as an example, the NDVI value calculation process is as follows: For pixel P2, its NDVI value is calculated as follows: After calculation, the obtained NDVI value is compared with the set vegetation index threshold. The threshold is set based on the training process of a deep learning model containing 10,000 labeled samples from multiple time periods and resolutions. The sample set has undergone image enhancement processing such as random rotation, translation, and addition of Gaussian noise to improve the robustness of the model. The model training uses a joint loss function of SoftCrossEntropyloss and Focalloss, and uses a cosine annealing strategy to adjust the learning rate. Through iterative optimization and performance evaluation, the threshold is determined to be set at 0.2. This threshold can achieve an overall classification accuracy of 97.8%, as shown in Table 1.

[0070] Table 1 Examples of Pixel NDVI Calculation and Determination

[0071] ;

[0072] As shown in Table 1, during the specific comparison, the NDVI value of pixel P1 is 0.733. Since 0.733 > 0.2, pixel P1 is marked as a vegetated area. The NDVI value of pixel P2 is 0.091. Since 0.091 < 0.2, P2 is marked as a non-vegetated area. This calculation and comparison process is repeated for all pixels along the transmission line. All pixels with NDVI values ​​higher than 0.2 are selected and marked as "1", while the remaining pixels are marked as "0", thus obtaining the vegetation index determination result.

[0073] The vegetation index determination results are called, and the spatial correspondence between the spatial coordinates of the pixels and the spatial data of the transmission lines is made. The intersection relationship between the vegetation area and the transmission line is detected, the spatial aggregation calculation is performed on the intersecting pixels, and the vegetation area corresponding to the transmission line is extracted based on the aggregated position coordinates to generate the remote sensing vegetation spatial distribution results.

[0074] The vegetation index determination results are retrieved using an overlapping sliding window prediction strategy. The original remote sensing image to be classified is cropped into 512×512 pixel image blocks with a cropping step of 384 pixels, resulting in a 128-pixel overlapping area. After predicting each image block separately, only the central 384×384 pixel area is selected, ignoring the 64-pixel edge area. The effective central areas are then stitched together to form a complete vegetation determination raster image. This raster image is post-processed to convert it from raster format to vector polygon format, and background polygons (determination value "0") are removed, retaining only vegetation (determination value "1") patches. The area of ​​each vegetation patch is then calculated, with an area threshold of 5 square meters. Small patches smaller than this threshold are removed, while holes smaller than 2 square meters within large patches are filled. The connectivity of disconnected but spatially adjacent patches is optimized. The processed vegetation area vector data is then compared with the transmission line spatial data. (Vector line files, recording the precise coordinates of towers and the route of the power transmission line) Spatial overlay analysis is performed to detect the spatial intersection relationship between the vegetation area polygons and the transmission line vector lines. Vegetation patches whose geometric centers fall within the 30-meter buffer zone of the transmission line are identified as intersecting pixels. For all identified intersecting pixels, spatial aggregation calculation is performed using the eight-neighborhood connection rule, merging consecutively adjacent intersecting pixels into an independent vegetation area polygon (e.g., Area001). Based on the aggregated location coordinates, the set of vegetation areas corresponding to the transmission line is extracted, generating remote sensing vegetation spatial distribution results. Regarding the automatic identification and labeling of vegetation areas, a sample library of potential safety hazards in transmission line corridors and a refined intelligent extraction model are constructed based on deep learning technology. For different targets, target detection (such as the YOLOV8 algorithm) and semantic segmentation (such as the DeeplabV3+ algorithm, and incorporating Squeeze-and-Excitation) are employed. Algorithms such as (SE) attention mechanism module are used for automated extraction. Among them, YOLOV8 is suitable for extracting features with regular boundaries such as trees, while DeeplabV3+ is suitable for detecting large-area targets with irregular boundaries, so as to improve extraction accuracy and robustness.

[0075] Please see Figure 3 The steps to obtain the safe distance area are as follows:

[0076] Extract vegetation area information from remote sensing vegetation spatial distribution results, obtain data on power transmission lines, including line direction and location, match these data with the geographic coordinates of vegetation areas, calculate the coordinate offset between power transmission lines and vegetation areas, correct the line direction, and generate a set of line spatial offset parameters.

[0077] The geometric center coordinates and boundary information of each aggregated vegetation polygon in the remote sensing vegetation spatial distribution results are extracted. For example, the center coordinates of Area001 are (116.318°, 39.925°). Simultaneously, the original vector data of the power transmission lines are acquired, and the vegetation area coordinates and power transmission line coordinates are matched to the same UTMZone50N projection coordinate system. The initial spatial relationship is determined by calculating the shortest straight-line distance from the boundary points of the vegetation area to the power transmission line segment. The coordinate offset between the power transmission line vector data and the remote sensing image is calculated. This offset is determined by selecting 50 clear ground features (such as road intersections) as ground control points and comparing them... The image coordinates and high-precision GPS measured coordinates were used to calculate the average offset in the east-west direction as +1.5 meters and the average offset in the north-south direction as -0.8 meters. Based on these offsets, each coordinate point in the transmission line vector data was corrected. For example, the original coordinate point (525350E, 4418200N) was corrected to (525351.5E, 4418199.2N). The corrected line direction more accurately reflects the actual situation. All the corrected coordinate point sequences, the original vegetation area coordinates, and the calculated offsets (Δx=+1.5m, Δy=-0.8m) were integrated to generate a line spatial offset parameter set.

[0078] Based on the line spatial offset parameter set, the relative position between the transmission line and the vegetation area is calculated, the shortest distance between the line and each vegetation area is measured, and these distances are compared with the set safety distance. Areas below the safety threshold are extracted to obtain the line vegetation distance coefficient set.

[0079] Based on the line spatial offset parameter set, the corrected transmission line coordinates and spatial location information of vegetation areas are retrieved to calculate the relative position between the transmission line and each vegetation area. For vegetation area Area001, all vertices on its boundary are traversed, and the Euclidean distance from each vertex to the corrected transmission line vector segment is calculated. The minimum value is selected as the shortest distance between the vegetation area and the line. For example, the shortest distance of Area001 The shortest distance for Area002 is 25.8 meters. The minimum distance is 15.3 meters. This minimum distance is compared to a set safety distance threshold, which is determined based on the safety clearance regulations for 220kV lines and wind deflection simulation results. This is the first-level safety distance threshold. The safety distance is set at 30 meters, with a secondary safety distance threshold of 20 meters. Area 001, at 25.8 meters, falls between 20 and 30 meters, falling within the potential risk range. Area 002, at 15.3 meters, is less than 20 meters, falling within the high-risk range. All areas with a distance less than the 30-meter safety threshold are extracted and assigned a distance-based coefficient. ,in The distance coefficient for Area001 is meters. The distance coefficient of Area002 is The identifier, shortest distance, and calculated distance coefficient of each vegetation area are summarized to obtain the set of vegetation distance coefficients for the route.

[0080] Based on the set of vegetation distance coefficients for the transmission line, and combined with the spatial location, vegetation type and vegetation density of the adjacent vegetation areas, the density rate and spatial overlap ratio of each neighborhood are calculated. The values ​​are compared with the distance coefficients of the transmission line to dynamically adjust the range of potential risk areas and generate safe distance areas.

[0081] Based on the vegetation distance coefficient set along the route, and combined with the spatial location, vegetation type, and vegetation density (measured by the average NDVI per unit area) of neighboring vegetation areas around each vegetation area, the scope of potential risk areas is dynamically adjusted. Specifically, for Area002 (distance coefficient 0.49), the vegetation areas within a 100-meter radius around it (such as Area003 and Area004) are analyzed, and the comprehensive density rate of this neighborhood is calculated, which is the sum of the areas of all vegetation areas in the neighborhood divided by the total area of ​​the neighborhood, resulting in a density rate of 0.65. At the same time, the spatial overlap ratio between Area002 and its neighboring areas is calculated, resulting in an overlap ratio of 0.35. The density rate of 0.65, the overlap ratio of 0.35, and the distance coefficient of 0.49 are weighted and summed, with the weights set as density rate 0.4, overlap ratio 0.3, and distance coefficient 0.3, to calculate the comprehensive risk index. The index is compared with a preset adjustment threshold of 0.5, which is determined based on the inversion analysis of historical treeline contradiction failure points. The original risk boundary of Area002 is expanded, with the expansion distance being equal to... Positive correlation, calculated as This dynamic adjustment calculation is performed on all potentially risky areas to generate safe distance zones.

[0082] Please see Figure 4 The steps to obtain the terrain correction dataset are as follows:

[0083] Based on the spatial location of the vegetation area within the safe distance, the topographic elevation data of the corresponding area is obtained, the elevation value of each pixel in the topographic data is extracted, the elevation data is matched with the spatial coordinates of the vegetation area, the elevation difference between the pixels in the vegetation area and the surrounding terrain is calculated, and a set of topographic elevation difference values ​​is generated.

[0084] Based on the spatial location of each vegetated area within the safe distance, a digital elevation model (DEM) is generated from the point cloud data of a LiDAR system with a resolution of 5 meters. The elevation value of each pixel overlapping with the vegetated area in the DEM data is extracted. For example, if the elevation value of vegetated pixel P3 is 152.4 meters, all non-vegetated pixels within a 10-meter radius outside the boundary of this vegetated area are extracted as a reference for the surrounding terrain, and their average elevation is calculated. The average elevation of the surrounding terrain is found to be 148.6 meters. The difference between the average elevation of pixel P3 within the vegetated area and the average elevation of the surrounding terrain is then calculated to obtain the elevation difference value. The positive value in meters reflects that the pixel position is higher than the surrounding ground and may represent the vegetation height. This elevation extraction and elevation difference calculation is repeated for all vegetation pixels located within the safe distance area. The coordinates, original elevation and calculated elevation difference value of each pixel are recorded to generate a set of terrain elevation difference values.

[0085] Based on the topographic elevation difference value set, the elevation difference of pixels in adjacent vegetation areas is compared, areas with elevation differences exceeding a set threshold are marked, and the elevation values ​​of the marked areas are corrected to generate an elevation correction dataset.

[0086] Based on a set of topographic elevation differences, the elevation differences between spatially adjacent vegetation area pixels are compared to identify and correct abnormal elevation values. For example, within Area002, the elevation difference of pixel P3 is 3.8 meters, while the elevation difference of its eight-neighbor pixel P4 is 15.1 meters, totaling 11.3 meters. This exceeds the set elevation change threshold of 3 meters, which is based on statistical analysis of the canopy morphology of typical tree species within the study area. Elevation changes between adjacent canopies of the same tree are typically within 3 meters. The difference between P3 and P4 is 11.3 meters, far exceeding this threshold. If the elevation difference exceeds the 3-meter threshold, the pixel P4 with the larger elevation difference value is marked as a region to be corrected. The elevation value of the marked region P4 is then corrected by replacing it with the mean of the elevation difference values ​​of the unmarked pixels in its eight neighborhoods. Assuming that the mean elevation difference value of the other 7 unmarked pixels in the neighborhood of P4 is 3.9 meters, the elevation difference value of P4 is corrected from 15.1 meters to 3.9 meters, and its original elevation value is adjusted accordingly. This marking and correction process is performed on all regions with elevation differences exceeding the threshold throughout the entire dataset to generate an elevation correction dataset.

[0087] Based on the elevation correction dataset, the elevation changes of the corrected area and the surrounding terrain are compared to identify the accuracy of the correction results. The corrected elevation data is then combined with vegetation area information to generate a terrain correction dataset.

[0088] After obtaining the terrain correction dataset, the elevation differences of different vegetation areas are classified and identified. Areas with similar elevation change trends, differences, and terrain characteristics are grouped into the same category and assigned corresponding elevation pattern labels. In the subsequent risk assessment process, the safety distance and risk range are adjusted based on the elevation pattern labels.

[0089] Based on the elevation correction dataset, the elevation changes of the corrected area and surrounding terrain are re-examined to assess the accuracy of the correction results. For example, the corrected elevation difference of pixel P4 is 3.9 meters, and its elevation difference with the neighboring pixel P3 (3.8 meters) is reduced to 0.1 meters, confirming the effectiveness of the correction. All corrected elevation data are combined with vegetation area information to generate a terrain correction dataset. Based on this dataset, the elevation differences of different vegetation areas are classified and labeled. The mean, standard deviation, and slope of the elevation differences of all pixels within each vegetation area (e.g., Area_002) are calculated. The average elevation difference is 4.5 meters, the standard deviation is 0.8 meters, and the average slope is 12 degrees. Based on these statistical characteristics, areas with similar elevation change trends, differences, and topographic features are grouped into the same category. For example, the parameters of Area002 (4.5 meters, 0.8 meters, 12 degrees) are compared with the parameter range of the "gentle slope arbor area" model (average elevation difference 3-8 meters, standard deviation less than 1.0 meter, average slope less than 15 degrees). Since they are completely consistent, it is assigned the elevation model label of "gentle slope arbor area". This label will be used to adjust the calculation weight of safety distance and risk interval in subsequent risk assessment.

[0090] Please see Figure 5 The steps for obtaining the results of the tree and transmission line clearance analysis are as follows:

[0091] Based on the spatial location and elevation data of vegetation areas in the terrain correction dataset, the changing trend of elevation values ​​of each vegetation cell is detected, the range of elevation difference of cells in the same area is determined, the set of cells with continuous elevation rise is identified and the center position of the set is determined, the center position is marked as the position of the tree crown apex, and the set of tree crown apex elevation values ​​is generated.

[0092] Based on the spatial location of vegetation areas and the corrected elevation data in the terrain-corrected dataset, the changing trend of vegetation cell elevation values ​​within each area is detected to identify individual trees and locate their crown apex. Specifically, a 3x3 moving window traverses all cells within the vegetation area, determining whether the elevation of the central cell is greater than the elevation values ​​of its eight neighboring cells. If the central cell's elevation value satisfies the condition of being greater than the elevation values ​​of all eight neighboring cells, then that central cell is identified as the crown apex of a tree. For example... For example, in Area002, the elevation value of cell P5 (116.3182°, 39.9251°) is 153.1 meters. The elevation values ​​of its eight neighboring cells are between 151.9 meters and 152.8 meters, all lower than 153.1 meters. Therefore, the position of P5 is marked as the crown vertex, and its elevation value of 153.1 meters is recorded. This local maximum detection algorithm is performed on the entire Area002 and all other vegetation areas to summarize the positions and elevation values ​​of all identified crown vertices and generate a set of crown vertex elevation values.

[0093] The tree canopy apex elevation value set and the lowest ground elevation value in the same area are called, the numerical difference between the two is calculated and recorded as the height value of each tree, the height sequence is established based on the tree spatial coordinates, the vegetation units with height differences exceeding the set threshold are filtered out, and the tree height difference sequence is generated.

[0094] The tree canopy vertex elevation set is retrieved, and combined with the lowest ground elevation value in the terrain correction dataset, the height of each identified tree is calculated. For the tree canopy vertex P5 identified in S401, its elevation is 153.1 meters. The elevation of the lowest ground point in its area is 148.2 meters (this ground elevation value is obtained by interpolation of ground points in LiDAR point cloud data). The difference between the two values ​​is calculated, which is the tree height. The height is recorded as meters, and a height sequence containing the tree ID, coordinates, and height is created based on the spatial coordinates of each tree. For example:

[0095] [(Tree01, P5coord, 4.9m), (Tree02, P6coord, 8.2m), (Tree03, P7coord, 5.1m)], filtering out vegetation units with height differences exceeding a set threshold. This threshold is based on the average annual growth height of the main tree species in the area, which is 1.5 meters. The height difference threshold is set to 3 meters to mark abnormally tall trees. In the above sequence, Tree02's height of 8.2 meters is... rice, The differences in height were significant, exceeding the 3-meter threshold, and therefore were specially marked. The height of all trees and their difference analysis results were integrated to generate a tree height difference sequence.

[0096] Based on the tree height difference sequence and the spatial coordinates of the transmission line, the vertical distance between the top of each tree crown and the lowest conductor point of the corresponding transmission line is calculated. The distance value is compared with the safety clearance standard to determine the tree number and the area where the distance is lower than the safety requirement, and the tree and transmission line clearance analysis results are generated.

[0097] After obtaining the results of the net clearance analysis of trees and power transmission lines, the net clearance distances of different vegetation areas are classified and labeled. Areas with similar net clearance distances, vegetation types and densities are grouped into the same category and assigned corresponding risk level labels. In subsequent risk assessments, differentiated risk control measures are implemented based on the risk level labels.

[0098] Based on the height and spatial coordinates of each tree in the tree height difference sequence, and retrieving the corrected three-dimensional spatial coordinates of the transmission line (including the conductor elevation to ground calculated based on the tower and conductor sag model), the vertical clearance distance between each tree crown apex and the corresponding lowest conductor point of the transmission line is calculated. As shown in Table 2:

[0099] Table 2 Examples of Tree Clearance Calculation and Safety Status Assessment

[0100] ;

[0101] Table 2 presents the clearance analysis results for some trees. For example, directly above Tree02, the elevation of the lowest conductor of the transmission line is 162.3 meters, and the elevation of its canopy apex is 155.5 meters. The calculated vertical clearance distance is... The distance value of 6.8 meters is compared with the minimum safe clearance distance standard of 7.0 meters for 220kV lines. Since 6.8 meters is less than 7.0 meters, Tree02 and its surrounding area Area002 are deemed to fail to meet safety requirements and their non-compliance status is recorded. This clearance distance calculation and comparison process is repeated for all identified trees to generate the clearance analysis results between trees and transmission lines. Based on this, areas with clearance distances between 6.5 and 7.0 meters, vegetation type of trees, and a density greater than 0.7 are uniformly labeled as "critical risk". This label will be used to implement differentiated risk control measures in subsequent risk assessments.

[0102] Please see Figure 6 The steps for obtaining the high-risk assessment results of transmission line trees are as follows:

[0103] Based on the clearance distance data in the tree and transmission line clearance analysis results, the clearance distance value of all pixels in each vegetation area is detected. The clearance distance of each pixel is compared item by item according to the set safety standard threshold. The difference between the clearance distance and the safety standard threshold is calculated. The clearance distance status in the area is judged based on the difference, and a clearance distance difference set is generated.

[0104] Based on the clearance distance data of each tree recorded in the tree and transmission line clearance analysis results, the clearance distance value of all tree pixels in each vegetation area is detected, and the clearance distance of each pixel is compared item by item according to the set safety standard threshold (7.0 meters). For example, for Area002, which contains two trees, Tree02 (clearance distance 6.8 meters) and Tree05 (clearance distance 7.5 meters), the difference between the clearance distance of each tree and the safety standard threshold is calculated. For Tree02, the difference Meters, for Tree05, the difference The distance is measured in meters. Negative values ​​indicate insufficient clearance, while positive values ​​indicate that the requirements are met. Based on the difference, the clearance status of each tree in the area is determined, and all calculated differences, along with their corresponding tree IDs and location coordinates, are stored to generate a clearance difference set.

[0105] The system retrieves numerical data from the set of airspace distance differences, filters out areas where the airspace distance is less than or equal to the minimum threshold of the safety standard based on the minimum threshold of the safety standard, and sets them as high-risk areas. It also classifies areas where the airspace distance difference is between the safety standard and the minimum threshold as medium-risk areas, and generates risk level division interval value groups.

[0106] The system retrieves numerical data from the set of clearance distance differences and classifies vegetation areas into risk levels based on a predefined risk classification standard. This standard sets a minimum safety threshold of 7.0 meters and a medium-risk buffer threshold of 8.0 meters. When the clearance distance difference is less than or equal to 0 (i.e., clearance distance less than or equal to 7.0 meters), the corresponding area is selected and classified as high-risk. For example, since the clearance distance difference for Tree02 is -0.2 meters, which is less than 0, its area Area002 is initially classified as high-risk. For risk levels, when the difference in airspace distance is greater than 0 but less than or equal to 1.0 meter (i.e., the airspace distance is between 7.0 and 8.0 meters), the corresponding area is classified as medium risk. If the airspace distance of Tree05 is 7.5 meters, the difference is +0.5 meters, then the area is classified as medium risk. The risk level division intervals, namely the high risk interval (-∞, 0), the medium risk interval (0, 1.0), and the subsequently defined low risk interval (1.0, +∞), are fixed to generate a risk level division interval value group.

[0107] Based on the remaining data in the risk level division interval value group and the clearance distance difference set, the area with the clearance distance greater than or equal to the safety standard threshold is determined. The risk level value of each area is merged with the corresponding spatial coordinates and area number. The risk value distribution is integrated according to the level interval to generate the high risk assessment result of the transmission line tree.

[0108] Based on the remaining data in the risk level interval value group and the set of clearance distance differences, areas with clearance distance differences greater than 1.0 meter (i.e., clearance distance greater than 8.0 meters) are determined to be low-risk areas. For example, assuming the clearance distance of Tree07 is 9.2 meters, its difference is +2.2 meters, which is greater than 1.0 meter, therefore its risk level is low. The risk level value determined for each area is combined with the spatial coordinates and area number of the corresponding area. For example, since Area002 contains the high-risk Tree02, the risk level of the entire area is determined to be high, recorded as (Area002, {coordinates}, high risk). After completing the level determination for all areas, GIS tools are used to automatically label and assign values ​​to the attribute fields of each risk point target, generating a structured risk point database, as shown in Table 3.

[0109] Table 3 Example of risk point database records

[0110] ;

[0111] Refer to Table 3, which lists a sample database record for a risk point target. This record includes a unique code, geographic coordinates, key risk parameters, and supplementary information. By integrating the numerical distribution of all risk points, a high-risk assessment result for the transmission line tree is generated. This result is synchronized to the field verification mobile application (APP) to support verification personnel in conducting on-site investigations. The APP has functions such as navigation, photo taking (limited to a photo location within a 50-meter radius of the map patch), information collection, and automatic result synchronization, realizing closed-loop management from automatic identification to accurate on-site verification.

[0112] A high-risk assessment device for power transmission line trees based on satellite remote sensing, comprising:

[0113] The first determination module is used to acquire the red and near-infrared bands of satellite remote sensing images, extract vegetation spectral reflectance values, identify vegetation pixels based on the difference in reflectance between the red and near-infrared bands, classify and determine vegetation types using vegetation indices, extract vegetation areas that intersect with power transmission lines by combining spatial coordinates, and generate remote sensing vegetation spatial distribution results.

[0114] The second determining module is used to determine the position and distance between the conductor and the vegetation based on the vegetation area information in the remote sensing vegetation distribution results, combined with the direction and location of the transmission line, to identify risk areas, dynamically adjust the range according to the characteristics of the neighboring vegetation, and generate a safe distance area.

[0115] The correction module is used to obtain the topographic elevation data corresponding to the vegetation area based on the spatial location of the vegetation within the safe distance area, analyze the elevation changes of the vegetation area and the surrounding terrain, compare the elevation difference of adjacent pixels, identify areas that exceed the threshold and correct them, and generate a terrain correction dataset.

[0116] The analysis module is used to identify the tree crown apex in the vegetation area based on the spatial location and elevation data of the vegetation area in the terrain correction dataset, calculate the tree height by comparing it with the ground elevation difference, obtain the clearance distance between the tree crown apex and the transmission line, identify trees or areas with insufficient clearance, and generate the tree and transmission line clearance analysis results.

[0117] A computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, implements the steps of any of the methods described above.

[0118] The above description is merely a specific embodiment of the present invention, but the scope of protection of the present invention is not limited thereto. Any variations or substitutions that can be easily conceived by those skilled in the art within the technical scope disclosed in the present invention should be included within the scope of protection of the present invention. Therefore, the scope of protection of the present invention should be determined by the scope of the claims.

Claims

1. A high-risk assessment method for power transmission line trees based on satellite remote sensing, characterized in that, Includes the following steps: The system acquires red and near-infrared bands from satellite remote sensing images, extracts vegetation spectral reflectance values, identifies vegetation pixels based on the difference in reflectance between red and near-infrared bands, classifies vegetation types using vegetation indices, extracts vegetation areas that intersect with power transmission lines using spatial coordinates, and generates remote sensing vegetation spatial distribution results. Based on the vegetation area information in the remote sensing vegetation spatial distribution results, spatial pairing is performed in conjunction with the transmission line route and location to determine the position and distance between the conductor and the vegetation, and to identify risk areas. Extract vegetation area information from the remote sensing vegetation spatial distribution results, obtain data on power transmission lines, including line direction and location, match these data with the geographic coordinates of the vegetation area, calculate the coordinate offset value between the power transmission line and the vegetation area, correct the line direction, and generate a set of line spatial offset parameters. Based on the set of line spatial offset parameters, the relative position of the transmission line and the vegetation area is calculated, the shortest distance between the line and each vegetation area is measured, and these distances are compared with the set safety distance. Areas below the safety threshold are extracted to obtain the set of line vegetation distance coefficients. Based on the set of vegetation distance coefficients for the transmission line, and combined with the spatial location, vegetation type and vegetation density of the adjacent vegetation areas, the density rate and spatial overlap ratio of each neighborhood are calculated. Then, the density rate, spatial overlap ratio and the distance coefficient of the transmission line are weighted to obtain a comprehensive risk coefficient, which is compared with a preset adjustment threshold. Based on the comparison results, the range of potential risk areas is dynamically adjusted to generate a safe distance area. Based on the spatial location of vegetation within the safe distance area, obtain the topographic elevation data corresponding to the vegetation area, analyze the elevation changes of the vegetation area and the surrounding terrain, compare the elevation differences of adjacent pixels, identify areas exceeding the threshold and correct them, and generate a terrain correction dataset. Based on the spatial location and elevation data of vegetation areas in the terrain correction dataset, the tree canopy apex of the vegetation area is identified, and the tree height is calculated by comparing it with the ground elevation difference. The clearance distance between the tree canopy apex and the power transmission line is obtained, trees or areas with insufficient clearance are identified, and the clearance analysis results between trees and power transmission lines are generated.

2. The high-risk assessment method for transmission line trees based on satellite remote sensing according to claim 1, characterized in that: The remote sensing vegetation spatial distribution results include vegetation spatial location, vegetation type classification, and information on areas where power transmission lines intersect. The safe distance area includes the spatial orientation of power transmission lines, relative distances between vegetation and risk zones. The terrain correction dataset includes vegetation area elevation values, terrain undulation parameters, and elevation correction coefficients. The tree and power transmission line clearance analysis results include tree height, crown elevation, and clearance distance data.

3. The high-risk assessment method for transmission line trees based on satellite remote sensing according to claim 1, characterized in that, The steps for obtaining the spatial distribution results of the remote sensing vegetation are as follows: The system acquires red and near-infrared reflectance data from satellite remote sensing images. It extracts red and near-infrared reflectance values ​​for each pixel in the image, calculates the difference between the two reflectance values ​​for the same pixel, classifies the pixels according to the set red and near-infrared difference threshold, selects a set of pixels that meet the vegetation characteristics, and generates a vegetation pixel reflectance difference set. Based on the vegetation pixel reflectance difference set, the red light to near-infrared reflectance ratio value is calculated for each pixel. The type is divided according to the ratio value and the set vegetation reflectance threshold range. The pixel distribution area of ​​different vegetation types is identified respectively. The coordinate positions of each type of pixel are spatially registered to generate a vegetation type distribution coefficient set. The set of vegetation type distribution coefficients is called, and the spatial coordinates of the pixels are spatially correlated based on the spatial coordinate data of the transmission lines. The intersection and proximity relationships are detected, and the intersection area is extracted based on the spatial distance threshold between the pixel and the transmission line. The relevant coordinates and vegetation type information are integrated to generate the spatial distribution results of remote sensing vegetation.

4. The high-risk assessment method for transmission line trees based on satellite remote sensing according to claim 1, characterized in that, The steps for obtaining the terrain correction dataset are as follows: Based on the spatial location of the vegetation area within the safe distance, the topographic elevation data of the corresponding area is obtained, the elevation value of each pixel in the topographic data is extracted, the elevation data is matched with the spatial coordinates of the vegetation area, the elevation difference between the pixels in the vegetation area and the surrounding terrain is calculated, and a set of topographic elevation difference values ​​is generated. Based on the terrain elevation difference value set, the elevation differences of adjacent vegetation area pixels are compared, areas with elevation differences exceeding a set threshold are marked, the elevation values ​​of the marked areas are corrected, and an elevation correction dataset is generated. Based on the elevation correction dataset, the elevation changes of the corrected area and the surrounding terrain are compared to identify the accuracy of the correction results. The corrected elevation data is then combined with vegetation area information to generate a terrain correction dataset.

5. The high-risk assessment method for transmission line trees based on satellite remote sensing according to claim 1, characterized in that, The steps for obtaining the clearance analysis results between trees and transmission lines are as follows: Based on the spatial location and elevation data of vegetation areas in the terrain correction dataset, the changing trend of elevation values ​​of each vegetation cell is detected, the range of elevation differences of cells in the same area is determined, the set of cells with continuously rising elevations is identified and the center position of the set is determined, the center position is marked as the position of the tree crown apex, and a set of tree crown apex elevation values ​​is generated. The tree canopy apex elevation value set and the lowest ground elevation value in the same area are called, the numerical difference between the two is calculated and recorded as the height value of each tree, a height sequence is established based on the tree spatial coordinates, vegetation units with height differences exceeding a set threshold are filtered out, and a tree height difference sequence is generated. Based on the tree height difference sequence and the spatial coordinates of the transmission line, the vertical distance between the top of each tree crown and the lowest conductor point of the corresponding transmission line is calculated. The distance value is compared with the safety clearance standard to determine the tree number and the area where the distance is lower than the safety requirement, and the tree and transmission line clearance analysis results are generated.

6. The high-risk assessment method for transmission line trees based on satellite remote sensing according to claim 1, characterized in that, The method further includes: Based on the clearance distance data in the tree and transmission line clearance analysis results, the clearance distance of each vegetation area is compared with the set safety standard. The risk level is divided according to the results: insufficient clearance is high risk, slightly below the standard is medium risk, and meeting the standard is low risk. The risk level and related data are combined to generate the high risk assessment result of the transmission line tree. The high-risk assessment results of the transmission line tree include risk level ranges, risk area distribution, and risk assessment indicators. The steps for obtaining the high-risk assessment results of the transmission line tree are as follows: Based on the clearance distance data in the clearance analysis results of trees and transmission lines, the clearance distance values ​​of all pixels in each vegetation area are detected. The clearance distance of each pixel is compared item by item according to the set safety standard threshold. The difference between the clearance distance and the safety standard threshold is calculated. The clearance distance status in the area is judged based on the difference, and a clearance distance difference set is generated. The numerical data in the set of airspace distance differences are called up. Based on the minimum threshold of the safety standard and the range of differences, areas with airspace distance less than or equal to the minimum threshold of the safety standard are selected and set as high-risk areas. Areas with airspace distance differences between the safety standard and the minimum threshold are classified as medium-risk areas, and risk level division interval value groups are generated. Based on the remaining data in the risk level division interval value group and the clearance distance difference set, the areas with clearance distance greater than or equal to the safety standard threshold are determined. The risk level values ​​of each area are merged with the corresponding spatial coordinates and area numbers, and the risk value distribution is integrated according to the level interval to generate the high risk assessment result of the transmission line tree.

7. The method for high-risk assessment of transmission line trees based on satellite remote sensing according to claim 1, characterized in that, After obtaining the terrain correction dataset, the elevation differences of different vegetation areas are classified and identified. Areas with similar elevation change trends, differences and terrain characteristics are grouped into the same group and assigned corresponding elevation pattern labels. In the subsequent risk assessment process, the safety distance and risk range are adjusted according to the elevation pattern labels. After obtaining the results of the clearance analysis of trees and power transmission lines, the clearance distances of different vegetation areas are classified and labeled. Areas with similar clearance distances, vegetation types and densities are grouped into the same category and assigned corresponding risk level labels. In subsequent risk assessments, differentiated risk control measures are implemented based on the risk level labels.

8. A high-risk assessment device for power transmission line trees based on satellite remote sensing, characterized in that, include: The first determination module is used to acquire the red and near-infrared bands of satellite remote sensing images, extract vegetation spectral reflectance values, identify vegetation pixels based on the difference in reflectance between the red and near-infrared bands, classify and determine vegetation types using vegetation indices, extract vegetation areas that intersect with power transmission lines by combining spatial coordinates, and generate remote sensing vegetation spatial distribution results. The second determining module is used to determine the position and distance between the conductor and the vegetation based on the vegetation area information in the remote sensing vegetation spatial distribution results, combined with the direction and location of the transmission line, and to identify risk areas. Extract vegetation area information from the remote sensing vegetation spatial distribution results, obtain data on power transmission lines, including line direction and location, match these data with the geographic coordinates of the vegetation area, calculate the coordinate offset value between the power transmission line and the vegetation area, correct the line direction, and generate a set of line spatial offset parameters. Based on the set of line spatial offset parameters, the relative position of the transmission line and the vegetation area is calculated, the shortest distance between the line and each vegetation area is measured, and these distances are compared with the set safety distance. Areas below the safety threshold are extracted to obtain the set of line vegetation distance coefficients. Based on the set of vegetation distance coefficients for the transmission line, and combined with the spatial location, vegetation type and vegetation density of the adjacent vegetation areas, the density rate and spatial overlap ratio of each neighborhood are calculated. Then, the density rate, spatial overlap ratio and the distance coefficient of the transmission line are weighted to obtain a comprehensive risk coefficient, which is compared with a preset adjustment threshold. Based on the comparison results, the range of potential risk areas is dynamically adjusted to generate a safe distance area. The correction module is used to obtain the topographic elevation data corresponding to the vegetation area based on the spatial location of the vegetation within the safe distance area, analyze the elevation changes of the vegetation area and the surrounding terrain, compare the elevation difference of adjacent pixels, identify areas that exceed the threshold and correct them, and generate a terrain correction dataset. The analysis module is used to identify the tree crown apex in the vegetation area based on the spatial location and elevation data of the vegetation area in the terrain correction dataset, calculate the tree height by comparing it with the ground elevation difference, obtain the clearance distance between the tree crown apex and the transmission line, identify trees or areas with insufficient clearance, and generate the tree and transmission line clearance analysis results.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that, When executed by a processor, the computer program implements the steps of the method described in any one of claims 1-7.

Images