Power transmission facility approach method and device based on remote sensing image recognition and electronic equipment

Abstract

Embodiments of the present disclosure disclose a power transmission facility approach method and device based on remote sensing image recognition and an electronic device. A specific embodiment of the method includes: acquiring a remote sensing image data set of a target power transmission channel; performing fusion correction processing on the remote sensing image data set to obtain a fused geographic information data set; performing ground object semantic segmentation to obtain a ground object element information set; performing vegetation point cloud segmentation processing to obtain a vegetation information set; performing crop type clustering to obtain a farmland plot vector set; determining building contour information and generating a building risk point information set; performing water area dynamic analysis to obtain a water area dynamic information set; generating a power transmission line vector layer set according to the vegetation information set, the farmland plot vector set, the building risk point information set, and the water area dynamic information set; and controlling a transportation device to complete the approach operation of the power transmission facility according to the power transmission line vector layer set. The embodiment can reduce the safety risk in the approach process of the power transmission facility.

Description

Technical Field

[0001] The embodiments of this disclosure relate to the field of computer technology, and more specifically to a method, apparatus, and electronic device for accessing power transmission facilities based on remote sensing image recognition. Background Technology

[0002] The on-site construction of power transmission facilities (such as towers, conductors, and heavy construction equipment) is a crucial part of the bookstore's power line project. Traditional methods for bringing power transmission facilities to the site mainly rely on the analysis of two-dimensional paper data, image data, or CAD drawings, combined with simple GIS tools to determine the access route in order to complete the on-site operation.

[0003] However, when using the above methods to control transportation facilities to complete the on-site operation of power transmission facilities, the following technical problems often arise: Existing solutions rely on static analysis from a single data source, which cannot overcome the interference of environmental factors. This results in a lack of risk identification for key features such as woodlands, farmland, buildings, and water bodies, making it difficult to meet planning needs under complex terrain and severe weather conditions. Consequently, there are safety risks associated with the installation of power transmission facilities.

[0004] The information disclosed in this background section is only intended to enhance the understanding of the background of the present disclosure concept, and therefore may contain information that does not constitute prior art known to those skilled in the art. Summary of the Invention

[0005] The summary portion of this disclosure is intended to provide a brief overview of the concepts, which will be described in detail in the detailed description portion. This summary portion is not intended to identify key or essential features of the claimed technical solutions, nor is it intended to limit the scope of the claimed technical solutions.

[0006] Some embodiments of this disclosure provide a method, apparatus, electronic device, and computer-readable medium for accessing power transmission facilities based on remote sensing image recognition, in order to solve one or more of the technical problems mentioned in the background section above.

[0007] In a first aspect, some embodiments of this disclosure provide a method for power transmission facility access based on remote sensing image recognition. The method includes: in response to receiving a power transmission facility access command, acquiring a remote sensing image dataset of the target power transmission channel through the aforementioned multi-source acquisition system; performing fusion correction processing on the aforementioned remote sensing image dataset to obtain a fused geographic information dataset; performing semantic segmentation of the aforementioned fused geographic information dataset using a remote sensing image segmentation model to obtain a set of geographic feature information; performing vegetation point cloud segmentation processing on the aforementioned geographic feature information set to obtain a vegetation information set; and performing crop type segmentation on the aforementioned geographic feature information set. Clustering is performed to obtain a vector set of farmland plots; the building outline information of the above-mentioned land feature information set is determined to obtain a building outline set, and a building risk point information set is generated based on the above-mentioned building outline set; water dynamic analysis is performed on the above-mentioned land feature information set to obtain a water dynamic information set; a power transmission line vector layer set is generated based on the above-mentioned vegetation information set, the above-mentioned farmland plot vector set, the above-mentioned building risk point information set, and the above-mentioned water dynamic information set; the power transmission facility access route is determined based on the above-mentioned power transmission facility access route, and the above-mentioned transportation equipment is controlled to complete the access operation of the above-mentioned power transmission facility based on the above-mentioned power transmission facility access route.

[0008] Secondly, some embodiments of this disclosure provide a power transmission facility access device based on remote sensing image recognition. The device includes: an acquisition unit configured to acquire a remote sensing image dataset of a target power transmission channel through the aforementioned multi-source acquisition system in response to receiving a power transmission facility access command; a fusion correction processing unit configured to perform fusion correction processing on the aforementioned remote sensing image dataset to obtain a fused geographic information dataset; a ground feature semantic segmentation unit configured to perform ground feature semantic segmentation on the aforementioned fused geographic information dataset using a remote sensing image segmentation model to obtain a ground feature element information set; a vegetation point cloud segmentation processing unit configured to perform vegetation point cloud segmentation processing on the aforementioned ground feature element information set to obtain a vegetation information set; and a crop type clustering unit configured to perform crop type clustering processing on the aforementioned ground feature element information set. The system performs crop type clustering on the information set to obtain a farmland plot vector set; a first determining unit is configured to determine the building outline information of the above-mentioned land feature information set to obtain a building outline set, and generate a building risk point information set based on the above-mentioned building outline set; a water dynamic analysis unit is configured to perform water dynamic analysis on the above-mentioned land feature information set to obtain a water dynamic information set; a generating unit is configured to generate a power transmission line vector layer set based on the above-mentioned vegetation information set, the above-mentioned farmland plot vector set, the above-mentioned building risk point information set, and the above-mentioned water dynamic information set; a second determining unit is configured to determine the power transmission facility access route based on the above-mentioned power transmission facility access route, and control the above-mentioned transportation equipment to complete the access operation of the above-mentioned power transmission facility based on the above-mentioned power transmission facility access route.

[0009] Thirdly, some embodiments of this disclosure provide an electronic device, including: one or more processors; and a storage device having one or more programs stored thereon, wherein when the one or more programs are executed by the one or more processors, the one or more processors implement the method described in any implementation of the first aspect above.

[0010] Fourthly, some embodiments of this disclosure provide a computer-readable medium having a computer program stored thereon, wherein the program, when executed by a processor, implements the method described in any of the implementations of the first aspect above.

[0011] The above embodiments of this disclosure have the following beneficial effects: the power transmission facility access method based on remote sensing image recognition in some embodiments of this disclosure can reduce the safety risks of power transmission facilities during the access process. Specifically, the reason for the safety risks of power transmission facilities during the access process is that existing solutions rely on static analysis of a single data source, which cannot overcome the interference of environmental factors, resulting in a lack of risk identification for key features such as forest land, farmland, buildings, and water bodies, and difficulty in coping with planning needs under complex terrain and severe weather conditions, thus leading to safety risks of power transmission facilities during the access process. Based on this, the power transmission facility access method based on remote sensing image recognition in some embodiments of this disclosure firstly, in response to receiving the power transmission facility access command, acquires a remote sensing image dataset of the target power transmission channel through the aforementioned multi-source acquisition system. Thus, multi-source data from multiple platforms can provide a high-quality data foundation for a comprehensive understanding of the site environment. Secondly, the aforementioned remote sensing image dataset is fused and corrected to obtain a fused geographic information dataset. Thus, the interference of environmental factors can be eliminated. Thirdly, the aforementioned fused geographic information dataset is semantically segmented using a remote sensing image segmentation model to obtain a set of geographic feature information. This allows for the identification and classification of ground features in the imagery. Next, vegetation point cloud segmentation is performed on the aforementioned ground feature information set to obtain a vegetation information set. This enables individual tree segmentation, allowing for accurate assessment of deforestation volume and providing a quantitative basis for planning construction routes. Subsequently, crop type clustering is performed on the aforementioned ground feature information set to obtain a farmland plot vector set. This allows for the differentiation of different crop types. Then, building outline information is determined from the aforementioned ground feature information set to obtain a building outline set, and based on this building outline set, a building risk point information set is generated. This allows for the assessment of safe distances between buildings and the construction route. Then, water dynamic analysis is performed on the aforementioned ground feature information set to obtain a water dynamic information set. This allows for the avoidance of flooded areas in the access route. Finally, based on the aforementioned vegetation information set, farmland plot vector set, building risk point information set, and water dynamic information set, a transmission line vector layer set is generated. This provides a directly operable decision-making environment for subsequent route planning. Finally, based on the aforementioned transmission line vector map layer set, the access route for the transmission facilities is determined, and the aforementioned transportation equipment is controlled to complete the access operation of the transmission facilities according to the access route. Ultimately, this reduces the safety risks associated with the access of transmission facilities during the process. Attached Figure Description

[0012] The above and other features, advantages, and aspects of the embodiments of this disclosure will become more apparent from the accompanying drawings and the following detailed description. Throughout the drawings, the same or similar reference numerals denote the same or similar elements. It should be understood that the drawings are schematic, and elements are not necessarily drawn to scale.

[0013] Figure 1 This is a flowchart of some embodiments of the power transmission facility access method based on remote sensing image recognition according to this disclosure; Figure 2 This is a schematic diagram of the structure of some embodiments of the power transmission facility access device based on remote sensing image recognition according to the present disclosure; Figure 3 This is a schematic diagram of the structure of an electronic device suitable for implementing some embodiments of the present disclosure; Figure 4 These are remote sensing image recognition result diagrams based on some embodiments of the power transmission facility access device based on remote sensing image recognition according to this disclosure. Detailed Implementation

[0014] Embodiments of this disclosure will now be described in more detail with reference to the accompanying drawings. While some embodiments of this disclosure are shown in the drawings, it should be understood that this disclosure can be implemented in various forms and should not be construed as limited to the embodiments set forth herein. Rather, these embodiments are provided to provide a more thorough and complete understanding of this disclosure. It should be understood that the accompanying drawings and embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of protection of this disclosure.

[0015] It should also be noted that, for ease of description, only the parts relevant to the invention are shown in the accompanying drawings. Unless otherwise specified, the embodiments and features described in this disclosure can be combined with each other.

[0016] It should be noted that the concepts of "first" and "second" mentioned in this disclosure are used only to distinguish different devices, modules or units, and are not used to limit the order of functions performed by these devices, modules or units or their interdependencies.

[0017] It should be noted that the terms "a" and "a plurality of" used in this disclosure are illustrative rather than restrictive, and those skilled in the art should understand that, unless otherwise expressly indicated in the context, they should be understood as "one or more".

[0018] The names of messages or information exchanged between multiple devices in the embodiments of this disclosure are for illustrative purposes only and are not intended to limit the scope of such messages or information.

[0019] This disclosure will now be described in detail with reference to the accompanying drawings and embodiments.

[0020] Figure 1A flowchart 100 is shown, illustrating some embodiments of a power transmission facility access method based on remote sensing image recognition according to this disclosure. This power transmission facility access method based on remote sensing image recognition is applied to a power transmission facility access system, which includes a multi-source acquisition system and transport equipment, and includes the following steps: Step 101: In response to receiving the power transmission facility access command, acquire the remote sensing image dataset of the target power transmission channel through the multi-source acquisition system.

[0021] In some embodiments, the execution entity (e.g., a server) of the power transmission facility access method based on remote sensing image recognition can, in response to receiving a power transmission facility access instruction, acquire a remote sensing image dataset of the target power transmission channel through the aforementioned multi-source acquisition system. The aforementioned power transmission facility access instruction can be a digital instruction to initiate the pre-construction preparation process for a specific power transmission line section. The aforementioned target power transmission channel can be the geographic spatial range traversed by the planned power transmission line. The aforementioned multi-source acquisition system can be a data acquisition and transmission system integrating multiple remote sensing data acquisition platforms (such as optical satellites, radar satellites, and UAV aerial photography systems). The remote sensing image data in the aforementioned remote sensing image dataset includes: visible light image data, thermal infrared image data, and point cloud data. The aforementioned power transmission facilities can include, but are not limited to, tower components (such as tower materials, tower feet, insulators), insulation equipment, and substation and support equipment (such as transformers, circuit breakers, and structural supports). The aforementioned transportation equipment can be vehicles or engineering machinery used to transport the aforementioned power transmission facilities. For example, it can include heavy-duty trucks, truck cranes, and excavators.

[0022] In practice, the aforementioned implementing entity may, in response to receiving the aforementioned power transmission facility entry instruction, send a data acquisition request to the aforementioned multi-source acquisition system to obtain the remote sensing image dataset of the aforementioned target power transmission channel.

[0023] Step 102: Perform fusion correction processing on the remote sensing image dataset to obtain a fused geographic information dataset.

[0024] In some embodiments, the aforementioned executing entity may perform fusion correction processing on the aforementioned remote sensing image dataset to obtain a fused geographic information dataset.

[0025] In some optional implementations of certain embodiments, the aforementioned executing entity may perform fusion correction processing on the aforementioned remote sensing image dataset through the following steps to obtain a fused geographic information dataset: Step one: Based on the aforementioned remote sensing image dataset, construct a multi-source image dataset. In practice, the executing entity can select at least one remote sensing image from the dataset that meets preset screening criteria, and then index and associate these images according to their data source and capture time to obtain the multi-source image dataset. The preset screening criteria can be selecting remote sensing image data whose capture time falls within the project cycle.

[0026] Step two involves performing geometric calibration and spatial registration on the aforementioned multi-source image dataset to obtain a registered image dataset. In practice, the executing entity can select the image data with the highest geometric accuracy from the multi-source image dataset as the reference image, extract the corresponding points of the reference image data and other image data in the multi-source image dataset, and transform the pixel coordinates of the other image data in the multi-source image dataset to the coordinate system of the reference image through polynomial transformation or affine transformation to obtain the registered image dataset.

[0027] Step three involves radiometric calibration and atmospheric correction of the registered image dataset to obtain a surface reflectance image set. In practice, the executing entity can perform radiometric calibration on the registered image dataset using the calibration coefficients provided in the multi-source acquisition system (sensor) to convert the original pixel values ​​into apparent reflectance at the top of the atmosphere. Then, using an atmospheric radiative transfer model (such as MODTRAN (moderate resolution transmission), 6S (second simulation of the satellite signal in the solar spectrum), or ACTOR (atmospheric and topographic correction)), combined with atmospheric parameters (such as water vapor content) during image imaging, atmospheric correction is performed on the apparent reflectance to obtain the surface reflectance image set. This surface reflectance image set can be a collection of image data that reflects the spectral characteristics of the ground features themselves.

[0028] Step four involves performing topographic radiometric correction on the aforementioned surface reflectance image set to obtain a topographically corrected image set. In practice, the implementing entity can determine the solar incidence angle corresponding to each pixel in each fused image set, and then correct the reflectance of each pixel in each fused image set according to the solar incidence angle corresponding to each pixel in each fused image set using a topographic radiometric correction model (such as an empirical model, SCS correction model, or C correction model), thereby obtaining the topographically corrected image set.

[0029] Step 5: Perform cloud removal and shadow compensation processing on the aforementioned terrain-corrected image set to obtain a cloud-free image set. In practice, the executing entity can use a thresholding method to detect clouds in the aforementioned surface reflectance image set according to the reflectance characteristics and texture features of clouds in specific bands, obtaining a cloud mask. Then, using spatiotemporal interpolation, perform regional restoration on the cloud mask to obtain a cloud-removed image set. Subsequently, detect areas with cloud shadows or terrain shadows in the aforementioned cloud-removed image set, and perform shadow compensation processing on these areas using grayscale linear mapping to obtain a cloud-free image set.

[0030] Step six involves performing image fusion processing on the aforementioned cloudless image set to obtain a fused image set. In practice, the aforementioned execution entity can use pansharpening technology to perform image fusion processing on the aforementioned cloudless image set to obtain a fused image set.

[0031] Step seven involves performing temporal filtering and feature enhancement on the aforementioned topographically corrected image set to obtain a fused geographic information dataset. In practice, the executing entity can use a temporal filtering algorithm (Savitzky-Golay filter) to perform temporal filtering on the topographically corrected image set to obtain a filtered image set. Then, the enhanced features of the filtered image set (such as normalized difference vegetation index, normalized difference water index, etc.) are determined to obtain the fused geographic information dataset.

[0032] Step 103: Using a remote sensing image segmentation model, perform semantic segmentation of the fused geographic information dataset to obtain a set of geographic feature information.

[0033] In some embodiments, the aforementioned executing entity can perform semantic segmentation of the fused geographic information dataset using a remote sensing image segmentation model to obtain a set of geographic feature information. In practice, the executing entity can input the fused information dataset into the remote sensing image segmentation model to obtain the set of geographic feature information. The geographic feature information in the set includes pixel outlines and corresponding geographic feature category attributes (such as vegetation, crops, buildings, and water bodies). The aforementioned remote sensing image segmentation model can be a deep learning model (such as Samba or RSAM-Seg) capable of classifying each pixel in the remote sensing image data. The aforementioned remote sensing image segmentation model can be trained on a large number of labeled remote sensing impact samples of power transmission corridors and can identify geographic feature categories including buildings, roads, farmland, and trees. The aforementioned remote sensing image segmentation model can include an input encoding and feature extraction layer, a context information aggregation layer, a feature encoding and resolution restoration layer, and a post-processing optimization layer. The aforementioned input encoding and feature extraction layer can take the fused geographic information dataset as input and a multi-scale feature atlas as output. The input encoding and feature extraction layer described above can be constructed using a deep convolutional neural network (such as ResNet). Through convolutional pooling operations, it extracts and abstracts multi-scale features from the input image data, ranging from low-level edge texture to high-level semantics. The context information aggregation layer described above can take the multi-scale feature map set as input and an enhanced feature map set as output. This layer can capture the semantic relationships between pixels in the multi-scale feature map set using an attention mechanism, resulting in an enhanced feature map set. The feature encoding and resolution restoration layer described above can take the enhanced feature map set as input and a preliminary pixel-level classification score map as output. This layer can fuse the high-level semantic features of the enhanced feature map set with the low-level detail features of the encoding layer through upsampling operations (such as transposed convolution or interpolation) and skip connections, gradually restoring the spatial size of the enhanced feature map and generating a probability for each pixel to belong to a land cover category. The post-processing optimization layer described above can take the preliminary pixel-level classification score map as input and a land cover feature information set as output. The aforementioned post-processing optimization layer uses the argmax operation to convert the initial pixel-level classification score map into a category label map. Then, through connectivity component analysis, it aggregates spatially connected pixels belonging to the same category into independent objects (patches) and determines the geometric contours (vector polygons) of each object, obtaining an initial set of ground feature information elements. Afterward, noisy patches are removed from the initial set of ground feature information elements, and the resulting set is smoothed to obtain the final ground feature information set. (See reference...) Figure 4 The remote sensing image recognition result diagrams are based on some embodiments of the power transmission facility access device based on remote sensing image recognition disclosed herein.

[0034] Step 104: Perform vegetation point cloud segmentation on the ground feature information set to obtain the vegetation information set.

[0035] In some embodiments, the aforementioned executing entity may perform vegetation point cloud segmentation processing on the aforementioned set of ground feature information to obtain a vegetation information set.

[0036] In some optional implementations of certain embodiments, the aforementioned execution entity may perform vegetation point cloud segmentation processing on the aforementioned land feature information set through the following steps to obtain the vegetation information set: Step one: Based on the above set of ground feature information, extract the patch extent of vegetation to obtain a vector information set of vegetation regions. In practice, the executing entity can extract the geometric boundaries of the ground feature information representing "vegetation" from the above set of ground feature information to obtain a vector information set of vegetation regions.

[0037] Step two: Obtain point cloud data that matches the aforementioned vegetation area vector information set to obtain the original point cloud dataset. In practice, the aforementioned execution entity can obtain point cloud data covering the spatial range of each vegetation area vector information in the aforementioned vegetation area vector information set from a pre-stored point cloud database generated by UAV measurements, thus obtaining the original point cloud dataset.

[0038] Step three involves normalizing the original point cloud dataset to obtain a normalized vegetation point cloud set. In practice, the execution entity can use a progressive triangulation algorithm to perform point cloud ground filtering on the original point cloud dataset, obtaining a ground point cloud set and a non-ground point cloud set. Then, each non-ground point cloud in the non-ground point cloud set is subtracted from its corresponding ground point cloud to obtain the normalized vegetation point cloud, thus obtaining the normalized vegetation point cloud set.

[0039] Step four involves segmenting the normalized vegetation point cloud set into individual tree point clouds using a single-tree segmentation algorithm, resulting in individual tree point cloud clusters. In practice, the executing entity can use the aforementioned single-tree segmentation algorithm to segment each normalized vegetation point cloud in the normalized vegetation point cloud set, obtaining individual tree point cloud clusters. The single-tree segmentation algorithm is a technique used to identify and separate individual trees from point cloud data. For example, it can be a CHM-based single-tree segmentation.

[0040] Step 5: Extract the geometric and spectral reflectance features of each individual tree point cloud cluster in the aforementioned individual tree point cloud cluster set to obtain an individual tree feature vector set. In practice, the executing entity can obtain the geometric feature set by determining the statistical distribution of the three-dimensional points of each individual tree point cloud cluster in the aforementioned individual tree point cloud cluster set. Then, determine the average reflectance of each individual tree point cloud cluster in the aforementioned individual tree point cloud cluster set to obtain the spectral reflectance feature set. Finally, combine the aforementioned geometric feature set and the aforementioned spectral reflectance feature set into an individual tree feature vector set.

[0041] Step six: Based on the aforementioned individual tree feature vector set, generate an individual tree species label set. In practice, the executing entity can use a tree species identification algorithm to identify the tree species of each individual tree feature vector in the aforementioned individual tree feature vector set, thereby obtaining the individual tree species label set. The aforementioned tree species identification algorithm can be a target detection algorithm based on a convolutional neural network (CNN), such as YOLO (You Only Look Once) and improved versions like YOLOv5 and YOLOv8.

[0042] Step seven: Generate a vegetation information set based on the aforementioned individual tree point cloud clusters and individual tree species tag sets. In practice, the executing entity can combine the aforementioned individual tree point cloud clusters and individual tree species tag sets into a vegetation information set.

[0043] Step 105: Cluster the land feature information set by crop type to obtain the farmland plot vector set.

[0044] In some embodiments, the executing entity can cluster the aforementioned set of geographic feature information by crop type to obtain a vector set of farmland plots. In practice, the executing entity can extract geographic feature information representing "farmland" from the aforementioned set of geographic feature information to obtain a set of farmland feature information. Then, using a clustering algorithm (such as K-means clustering), the farmland feature information set is clustered to obtain a set of crop type clusters. Finally, the farmland feature information set is bound to the category labels of the corresponding crop type clusters in the aforementioned crop type clusters to obtain a vector set of farmland plots.

[0045] Step 106: Determine the building outline information of the land feature information set to obtain the building outline set, and generate the building risk point information set based on the building outline set.

[0046] In some embodiments, the execution entity may determine the building outline information of the above-mentioned feature information set to obtain a building outline set, and generate a building risk point information set based on the above-mentioned building outline set.

[0047] In addressing the aforementioned technical challenges by employing technical solutions, the application scenario—transmission lines traversing densely populated and irregularly distributed building areas (such as urban villages and rural settlements)—often presents the following technical problems: traditional image recognition methods struggle to segment complete building outlines in densely built, overlapping areas. This makes it impossible to identify safe distances between the approach route and high-risk buildings during subsequent path planning, leading to collisions and damage to the transmission facilities during their approach. Considering the following requirements for this application scenario—quantifiable building outlines, distinguishable building types, differentiated risk assessment, and quantifiable construction safety distances—we have decided to adopt the following solution: In some optional implementations of certain embodiments, the execution entity may determine the building outline information of the aforementioned feature information set through the following steps to obtain a building outline set, and generate a building risk point information set based on the aforementioned building outline set: Step 1: Extract building pixels from the aforementioned feature information set to obtain a building binary mask. In practice, the execution entity can extract the feature information representing "buildings" from the aforementioned feature information set to obtain a building feature information set. Then, a blank raster layer with the same spatial range and resolution as the original image can be created, and the positions corresponding to the aforementioned building feature information set can be assigned a value of 1, while the remaining positions can be assigned a value of 0, thus obtaining the building binary mask.

[0048] Step two involves contour tracking of the aforementioned binary building mask to obtain an initial building contour set. In practice, the execution entity can use a contour tracking algorithm to perform contour tracking on the aforementioned binary building mask to obtain the initial building contour set. The contour tracking algorithm can be the Suzuki 85 algorithm. Step three involves regularizing the initial building outline set to obtain a building outline set. In practice, the execution entity can use the Douglas-Peucker algorithm to thin the vertices of each initial building outline in the initial building outline set, resulting in a thinned building outline set. Then, using a right-angled polygon fitting algorithm based on the minimum bounding rectangle, the thinned building outlines in the thinned building outline set are adjusted into regular polygons composed of line segments, thus obtaining the building outline set.

[0049] Step four: Perform connected component analysis on the above building outline set to obtain an independent building outline set. In practice, the above-mentioned execution entity can merge building outlines in the above building outline set whose edge distance is less than a preset threshold (e.g., 0.5 meters) and belong to the same connected region into an independent building outline, and divide the building outline set whose area exceeds a preset area threshold (e.g., 500 square meters) into multiple independent building outlines to obtain an independent building outline set.

[0050] Step 5: Based on the aforementioned remote sensing image dataset, perform multi-dimensional feature extraction on the aforementioned independent building outline set to obtain a multi-dimensional feature vector set for buildings. In practice, the executing entity can extract remote sensing image data corresponding to the aforementioned independent building outline set from the aforementioned remote sensing image dataset to obtain a building image dataset. Then, determine the average spectral value and texture feature statistics of each building image data in the aforementioned building image dataset to obtain a multi-dimensional feature vector set for buildings.

[0051] Step six involves performing type recognition on the aforementioned multi-dimensional building feature vector set to obtain a building type label set. In practice, the building type label can be identified for each multi-dimensional feature vector in the aforementioned multi-dimensional building feature vector set. As an example, a building type recognition algorithm can be used to identify the building type label for each multi-dimensional feature vector in the aforementioned multi-dimensional building feature vector set. This building type recognition algorithm can be the Mask R-CNN algorithm, which is suitable for building extraction and type recognition.

[0052] Step seven: Based on the aforementioned set of independent building outlines, generate a set of building geometric parameters. In practice, the executing entity can determine the length, width, area, and perimeter of each independent building outline in the set of independent building outlines to obtain the set of building geometric parameters.

[0053] Step 8: Based on the aforementioned set of building geometric parameters, the aforementioned set of building type labels, and the aforementioned set of independent building outlines, determine the building risk point information set. In practice, the executing entity can associate the aforementioned set of building geometric parameters, the aforementioned set of building type labels, and the aforementioned set of independent building outlines representing the same building to obtain a building information set. Then, determine the minimum distance between each building in the aforementioned building information set and the aforementioned target power transmission channel to obtain a minimum distance set, and based on the aforementioned minimum distance set, determine the building risk level to obtain the building risk point information set. As an example, buildings with a minimum distance of less than 3 meters can be identified as high-risk. Buildings with a minimum distance greater than 3 meters but less than 10 meters can be identified as medium-risk. Buildings with a minimum distance greater than 10 meters can be identified as high-risk.

[0054] Steps one through eight and their related content, as an inventive point of this disclosure, combined with step "109" below, solve the technical problem that "traditional image recognition methods struggle to segment complete building outlines in densely built, overlapping areas, making it impossible to identify safe distances between the approach route and high-risk buildings in subsequent path planning, leading to collisions with power transmission facilities during the approach process and consequently, damage to the power transmission facilities." The reason for the damage to power transmission facilities is that traditional image recognition methods struggle to segment complete building outlines in densely built, overlapping areas, making it impossible to identify safe distances between the approach route and high-risk buildings in subsequent path planning, leading to collisions with power transmission facilities during the approach process and consequently, damage to the power transmission facilities. Solving these factors can resolve the problem of power transmission facility damage. To achieve this, the first step is to extract building pixels from the aforementioned set of ground feature information to obtain a binary building mask. This allows for the initial identification of building targets from a cluttered scene. The second step is to perform contour tracking on the aforementioned binary building mask to obtain an initial set of building outlines. This allows the building outlines to be converted into vector boundaries. The third step involves regularizing the initial building outline set to obtain a building outline set. This eliminates noise and distortion caused by image segmentation. The fourth step involves performing connected component analysis on the building outline set to obtain an independent building outline set. This allows for the identification of individual building entities. The fifth step involves extracting multi-dimensional features from the independent building outline set based on the remote sensing image dataset, resulting in a multi-dimensional feature vector set for buildings. This provides quantitative data for distinguishing building materials and structures. The sixth step involves type identification from the multi-dimensional feature vector set for buildings, resulting in a building type label set. This allows for the identification of building types. The seventh step involves generating a building geometric parameter set based on the independent building outline set. This provides the basic data for building outlines. The eighth step involves determining a building risk point information set based on the building geometric parameter set, the building type label set, and the independent building outline set. This allows for the determination of safe distances between access routes and high-risk buildings. In conjunction with step 109 below, the approach route of the power transmission facility is determined based on the aforementioned power transmission line vector map layer set. Then, based on this approach route, the transport equipment is controlled to complete the approach operation of the power transmission facility. Ultimately, damage to the power transmission facility can be avoided.

[0055] Step 107: Perform water area dynamic analysis on the land feature information set to obtain the water area dynamic information set.

[0056] In some embodiments, the aforementioned implementing entity may perform water area dynamic analysis on the aforementioned set of land feature information to obtain a set of water area dynamic information.

[0057] In addressing the aforementioned technical challenges by employing technical solutions, the application scenario—transmission lines traversing dynamically changing water areas such as seasonal rivers and lakes—often presents the following technical problems: traditional water area identification methods cannot dynamically capture the changing patterns of water bodies with seasonality and extreme weather, leading to the placement of planned access lines in areas exposed during the dry season or submerged during the rainy season, thus posing safety risks to transmission facilities and personnel. Considering the following requirements for this application scenario—adaptability to seasonal changes and predictable risks—we have decided to adopt the following solution: In some optional implementations of certain embodiments, the aforementioned executing entity may perform water area dynamic analysis on the aforementioned set of geographic feature information through the following steps to obtain a set of water area dynamic information: Step one: Based on the aforementioned set of geographic feature information, generate an initial water area vector set. In practice, the executing entity can extract the geographic feature information representing "water area" from the aforementioned set of geographic feature information and extract geometric features to obtain the initial water area vector set.

[0058] Step two: Obtain the historical water area image dataset corresponding to the initial water area vector set within a preset time window. The preset time window can be 3 years. The historical water area image data in the dataset can be remote sensing image data covering the spatial orientation of each initial water area vector in the initial water area vector set.

[0059] Step 3: Based on the water body index method, water bodies are extracted from the aforementioned historical water area image dataset to obtain a historical water area distribution sequence. In practice, the executing entity can determine the water body index of the aforementioned historical water area image dataset using the water body index method, and identify historical water area image data with a water body index greater than a preset water body threshold (e.g., 0.1) as water body images to generate water body masks, thus obtaining a water body mask set. Then, each water body mask in the aforementioned water body mask set undergoes denoising and small patch removal processing to obtain a water body vector dataset. Finally, the aforementioned water body vector dataset is sorted in ascending order according to time sequence to obtain the historical water area distribution sequence.

[0060] Step four involves performing spatiotemporal variation analysis on the aforementioned historical water body distribution sequence to obtain a water inundation frequency map and a dataset of variation patterns. In practice, the executing entity can rasterize the vector data of each water body in the aforementioned historical water body distribution sequence into the same grid. Then, the frequency at which each pixel in the grid is labeled as a water body is calculated to obtain the water inundation frequency map. Finally, the curves showing the change in water body area over time in the aforementioned grid are calculated, and the peak values ​​(flood season) and trough values ​​(dry season) of water body area are identified to obtain a dataset of variation patterns.

[0061] Step 5: Obtain the synthetic aperture radar (SAR) dataset corresponding to the initial water area vector set. In practice, the execution entity can obtain the SAR data for the spatial range corresponding to each initial water area vector in the initial water area vector set, thus obtaining the SAR dataset.

[0062] Step six: Based on the aforementioned synthetic aperture radar (SAR) dataset, generate a flood simulation result set. In practice, the aforementioned execution entity can use a hydrodynamic model to simulate the flood inundation range of each SAR data point in the dataset, thereby obtaining the flood simulation result set.

[0063] Step 7: Based on the aforementioned historical water area distribution sequence, the aforementioned water area inundation frequency map, the aforementioned variation pattern dataset, and the aforementioned flood simulation result set, a spatial overlay analysis is performed to obtain a water area risk level classification map. In practice, the implementing entity can spatially register and overlay the aforementioned historical water area distribution sequence, the aforementioned water area inundation frequency map, the aforementioned variation pattern dataset, and the aforementioned flood simulation result set to obtain a water area spatial information set. Then, the aforementioned water area spatial information set is classified to obtain a water area risk level classification map. As an example, areas with an inundation frequency greater than 70% can be classified as high-risk areas. Areas with an inundation frequency between 10% and 70% can be classified as medium-risk areas. Areas with an inundation frequency less than 10% can be classified as low-risk areas.

[0064] Step 8: Based on the aforementioned water area risk level classification map, the aforementioned change pattern dataset, and the aforementioned flood simulation result set, generate a water area dynamic information set. In practice, the implementing entity can associate the aforementioned water area risk level classification map, the aforementioned change pattern dataset, and the aforementioned flood simulation result set to obtain the water area dynamic information set.

[0065] Steps one through eight and their related content, as an inventive point of this disclosure, combined with step 109 below, solve the technical problem that "traditional water area identification methods cannot dynamically capture the dynamic changes of water areas with seasons and extreme weather, leading to the placement of planned access lines in areas exposed during the dry season or submerged during the rainy season, thus posing safety risks to power transmission facilities and personnel." The reason for these safety risks is that traditional water area identification methods cannot dynamically capture the dynamic changes of water areas with seasons and extreme weather, leading to the placement of planned access lines in areas exposed during the dry season or submerged during the rainy season, thus posing safety risks to power transmission facilities and personnel. Solving these factors can resolve the safety risks to power transmission facilities and personnel. To achieve this, the first step is to generate an initial water area vector set based on the aforementioned set of geographic feature information. This allows the determination of the water area's extent and spatial reference. The second step is to obtain a historical water area image dataset corresponding to the initial water area vector set within a preset time window. This allows the acquisition of historical water area images. The third step is to extract water bodies from the historical water area image dataset based on the water body index method, obtaining a historical water area distribution sequence. Therefore, multi-temporal image data can be transformed into structured information for spatiotemporal analysis. The fourth step involves performing spatiotemporal variation analysis on the aforementioned historical water area distribution sequence to obtain a water inundation frequency map and variation pattern dataset. This provides information on water area changes. The fifth step involves acquiring the synthetic aperture radar dataset corresponding to the initial water area vector set. This ensures the continuity and reliability of the data. The sixth step involves generating a flood simulation result set based on the aforementioned synthetic aperture radar dataset. This allows for the identification of high-risk areas. The seventh step involves performing spatial overlay analysis on the aforementioned historical water area distribution sequence, the aforementioned water inundation frequency map, the aforementioned variation pattern dataset, and the aforementioned flood simulation result set to obtain a water area risk level classification map. This transforms complex data into intuitive risk level information. The eighth step involves generating a water area dynamic information set based on the aforementioned water area risk level classification map, the aforementioned variation pattern dataset, and the aforementioned flood simulation result set. This provides reliable data for route planning. Combined with step "109" below, the access route for the transmission facilities is determined based on the aforementioned transmission line vector layer set, and the aforementioned transportation equipment is controlled to complete the access operation of the aforementioned transmission facilities based on the aforementioned access route. Ultimately, this can resolve the safety risks faced by power transmission facilities and personnel.

[0066] Step 108: Generate a vector layer set of transmission lines based on the vegetation information set, farmland plot vector set, building risk point information set, and water area dynamic information set.

[0067] In some embodiments, the execution entity may generate a vector layer set of transmission lines based on the vegetation information set, the farmland plot vector set, the building risk point information set, and the water area dynamic information set.

[0068] In some optional implementations of certain embodiments, the aforementioned execution entity may generate a transmission line vector layer set based on the aforementioned vegetation information set, the aforementioned farmland plot vector set, the aforementioned building risk point information set, and the aforementioned water area dynamic information set through the following steps: Step one involves performing coordinate system transformation on the aforementioned vegetation information set, farmland plot vector set, building risk point information set, and water area dynamic information set to obtain a multi-source dataset. In practice, the implementing entity can use a Geographic Information System (GIS) to transform the aforementioned vegetation information set, farmland plot vector set, building risk point information set, and water area dynamic information set to a unified coordinate system (such as the CGCS2000 coordinate system or the WGS84 coordinate system) to obtain a multi-source dataset.

[0069] Step two involves performing 3D modeling on the aforementioned multi-source dataset to obtain a 3D power transmission scene model of the target power transmission channel. In practice, the executing entity can utilize the acquired Digital Surface Model (DSM) to convert the data in the multi-source dataset into a geographic entity model to generate a 3D power transmission scene model of the target power transmission channel. This 3D power transmission scene model of the target power transmission channel can be a 3D digital model including 3D terrain and vegetation, farmland, buildings, and water bodies.

[0070] Step 3: Encapsulate the above-mentioned 3D power transmission scenario model to obtain a 3D geographic entity set. In practice, the aforementioned execution entity can perform lightweight processing on the above-mentioned 3D scenario model and encapsulate the classification code and attributes (such as building risk level and water inundation frequency) of each feature in the above-mentioned 3D power transmission scenario model to obtain a 3D geographic entity set.

[0071] Step four involves spatially associating the aforementioned 3D geographic entity set to obtain a transmission line vector layer set. In practice, the executing entity can determine the spatial relationships (such as distance and intersection angle) between various 3D geographic entities in the aforementioned 3D transmission scene model, thus obtaining a spatial relationship set. Then, the spatial relationship set, the aforementioned 3D geographic entity set, and the aforementioned 3D transmission scene model are packaged according to a geographic information data format (such as Shapefile) to obtain the transmission line vector layer set.

[0072] Step 109: Determine the access route of the power transmission facilities based on the vector map layer set of the power transmission lines, and control the transportation equipment to complete the access operation of the power transmission facilities based on the access route of the power transmission facilities.

[0073] In some embodiments, the execution entity may determine the access route of the power transmission facility based on the aforementioned power transmission line vector map layer set, and control the transportation equipment to complete the access operation of the power transmission facility based on the aforementioned access route of the power transmission facility.

[0074] In addressing the aforementioned technical problems in the process of adopting technical solutions, and considering the application scenario—transporting heavy power transmission facilities to complete on-site operations in complex terrain environments—the following technical issues often arise: Traditional path planning algorithms, based on two-dimensional maps or simple rules, cannot quantify the spatial conflicts and safe distances between construction equipment and complex three-dimensional terrain, nor can they predict the feasibility and risks of the path. This leads to the generated access routes being susceptible to collisions and personnel injuries due to equipment and terrain constraints during actual execution, thus posing safety risks during transportation. Considering the following requirements for this application scenario: adaptability to dynamic environments and collision simulation, we decided to adopt the following solution: In some optional implementations of certain embodiments, the execution entity may determine the access route of the power transmission facility based on the aforementioned power transmission line vector map layer set through the following steps, and control the transportation equipment to complete the access operation of the power transmission facility based on the aforementioned access route of the power transmission facility: Step one: Based on the aforementioned transmission line vector map layer set, generate the approach constraint parameter set. In practice, the executing entity can extract the constraint information from the aforementioned transmission line vector map layer set to obtain the approach constraint parameter set. The constraint information in this constraint information set can be parameter information used to restrict construction path planning. The constraint information in this constraint information set can include terrain slope, surface cover type (farmland, buildings, etc.), minimum safe distance between features and the approach line, bridge height limits, road weight limits, etc.

[0075] Step two: Based on the aforementioned set of entry constraint parameters and the pre-set construction equipment model library, a digital twin of the construction scene is constructed. In practice, the executing entity can extract equipment models related to the aforementioned transportation equipment from the pre-set construction equipment model library to obtain an equipment model set. Then, using the various entry constraint parameters in the aforementioned set of entry constraint parameters as constraints, a virtual model of the construction scene is constructed in a 3D simulation environment, and the aforementioned equipment model set is added to the virtual model to obtain the digital twin of the construction scene. The aforementioned pre-set construction equipment model library can be a database storing various construction transportation vehicle and equipment models and their corresponding kinematic parameters (such as dimensions, weight, minimum turning radius, etc.). The aforementioned digital twin of the construction scene can be a dynamic simulation model created in a virtual space (such as Unreal Engine) to reflect the actual construction site terrain, terrain constraints, and physical characteristics of the construction equipment.

[0076] Step 3: Based on the aforementioned digital twin of the construction scenario, determine a preliminary set of path schemes. In practice, the aforementioned execution entity can use a path planning algorithm (such as the A* algorithm) to generate at least one preliminary path scheme from the aforementioned digital twin of the construction scenario, thus obtaining a preliminary set of path schemes.

[0077] Step four involves performing path simulation on the aforementioned preliminary path scheme set to obtain a simulation result set. In practice, the executing entity can obtain the simulation result set by simulating the movement trajectory of each preliminary path scheme in the aforementioned preliminary path scheme set within the digital twin of the construction scenario. The simulation results in this set can include information such as the status of the construction equipment at each location, the degree of collision (no collision, minor collision, moderate collision, and severe collision), ground bearing capacity, and passage time.

[0078] Step 5: Based on the above simulation results set, fine-tune the feasibility of the preliminary path scheme set to obtain a feasible path scheme set. In practice, the executing entity can identify the preliminary path schemes in the above preliminary path scheme set that represent no collisions as feasible path schemes. Furthermore, for the preliminary path schemes in the above preliminary path scheme set that represent minor collisions, a local path replanning algorithm can be used to replan the road segments where minor collisions occur, obtaining feasible path schemes to generate a feasible path scheme set.

[0079] Step Six: Based on real-time environmental data, quantify the risks of the aforementioned set of feasible path options to obtain a path risk level set. In practice, the implementing entity can extract the level labels of the real-time environmental data corresponding to each feasible path option in the set to determine the path risk level and obtain the path risk level set. The aforementioned real-time environmental data can characterize the environmental data of the target transmission channel, such as real-time wind speed, rainfall, and soil moisture. For example, in response to the aforementioned real-time environmental data indicating a real-time wind speed less than 10 m / s, rainfall less than 10 mm, and soil moisture less than 25%, a path risk level representing low risk is generated. In response to the aforementioned real-time environmental data indicating a real-time wind speed greater than 10 m / s but less than 15 m / s, rainfall greater than 10 mm but less than 25 mm, and soil moisture greater than 25% but less than 35%, a path risk level representing medium risk is generated. In response to the aforementioned real-time environmental data indicating a real-time wind speed greater than 15 m / s, rainfall greater than 25 mm, and soil moisture greater than 35%, a path risk level representing low risk is generated.

[0080] Step seven involves weight optimization of the aforementioned path risk level set to obtain a path weight result set. In practice, the implementing entity can use multi-criteria decision-making (MCDM) to assign weights to the multi-dimensional indicators in the path risk level set and perform weighted summation to determine the path weight result set. These multi-dimensional indicators may include, but are not limited to, total path length, comprehensive safety risk, and estimated engineering costs (such as earthwork volume, crop compensation costs, and timber felling volume).

[0081] Step 8: Based on the above path weight result set, determine the access route for the transmission facilities. In practice, the implementing entity can determine the feasible path scheme corresponding to the path weight result with the highest weight in the above path weight result set as the access route for the transmission facilities.

[0082] Step nine: Based on the aforementioned power transmission facility access route, control the aforementioned transport equipment to complete the access operation of the aforementioned power transmission facility. In practice, the aforementioned executing entity can convert the aforementioned power transmission facility access route into machine-executable instructions to control the aforementioned transport equipment to complete the access operation of the aforementioned power transmission facility.

[0083] Steps one through nine and their related content, as an inventive point of this disclosure, solve the technical problem that "traditional path planning algorithms, based on two-dimensional maps or simple rules, cannot quantify the spatial conflicts and safety distances between construction equipment and complex three-dimensional terrain, and cannot predict the feasibility and risks of the path. This leads to the generated access routes being easily constrained by equipment and terrain during actual execution, causing collisions and personnel injuries, thus resulting in safety risks during transportation." The reason for these safety risks is that traditional path planning algorithms, based on two-dimensional maps or simple rules, cannot quantify the spatial conflicts and safety distances between construction equipment and complex three-dimensional terrain, and cannot predict the feasibility and risks of the path. This leads to the generated access routes being easily constrained by equipment and terrain during actual execution, causing collisions and personnel injuries, thus resulting in safety risks during transportation. Solving these factors can resolve the safety risks during transportation. To achieve this, the first step is to generate an access constraint parameter set based on the aforementioned power transmission line vector layer set. This allows the three-dimensional geographic information to be converted into machine-recognizable parameter information. The second step is to construct a digital twin of the construction scene based on the aforementioned access constraint parameter set and a preset construction equipment model library. This allows the construction of a virtual construction environment. The third step is to determine a preliminary set of path schemes based on the digital twin of the construction scenario. This allows for the formation of multiple alternative schemes. The fourth step is to perform path simulation on the preliminary set of schemes to obtain a simulation result set. This allows for the detection of collisions between equipment and objects in the environment. The fifth step is to fine-tune the preliminary set of path schemes based on the simulation result set to obtain a set of feasible path schemes. This allows for the elimination of routes with severe collision risks. The sixth step is to quantify the risk of the feasible path scheme set based on real-time environmental data to obtain a set of path risk levels. This allows for quantitative evaluation of routes in conjunction with real-time environmental data. The seventh step is to optimize the weights of the path risk level set to obtain a set of path weights. This allows for the quantification of path selection results. The eighth step is to determine the access route for the power transmission facilities based on the path weight result set. This allows for the selection of the optimal access route. The ninth step is to control the transportation equipment to complete the access operation of the power transmission facilities based on the access route. This allows the path to be converted into machine-executable instructions, directly guiding the equipment to execute them. Ultimately, this can reduce safety risks during transportation.

[0084] The above embodiments of this disclosure have the following beneficial effects: the power transmission facility access method based on remote sensing image recognition in some embodiments of this disclosure can reduce the safety risks of power transmission facilities during the access process. Specifically, the reason for the safety risks of power transmission facilities during the access process is that existing solutions rely on static analysis of a single data source, which cannot overcome the interference of environmental factors, resulting in a lack of risk identification for key features such as forest land, farmland, buildings, and water bodies, and difficulty in coping with planning needs under complex terrain and severe weather conditions, thus leading to safety risks of power transmission facilities during the access process. Based on this, the power transmission facility access method based on remote sensing image recognition in some embodiments of this disclosure firstly, in response to receiving the power transmission facility access command, acquires a remote sensing image dataset of the target power transmission channel through the aforementioned multi-source acquisition system. Thus, multi-source data from multiple platforms can provide a high-quality data foundation for a comprehensive understanding of the site environment. Secondly, the aforementioned remote sensing image dataset is fused and corrected to obtain a fused geographic information dataset. Thus, the interference of environmental factors can be eliminated. Thirdly, the aforementioned fused geographic information dataset is semantically segmented using a remote sensing image segmentation model to obtain a set of geographic feature information. This allows for the identification and classification of ground features in the imagery. Next, vegetation point cloud segmentation is performed on the aforementioned ground feature information set to obtain a vegetation information set. This enables individual tree segmentation, allowing for accurate assessment of deforestation volume and providing a quantitative basis for planning construction routes. Subsequently, crop type clustering is performed on the aforementioned ground feature information set to obtain a farmland plot vector set. This allows for the differentiation of different crop types. Then, building outline information is determined from the aforementioned ground feature information set to obtain a building outline set, and based on this building outline set, a building risk point information set is generated. This allows for the assessment of safe distances between buildings and the construction route. Then, water dynamic analysis is performed on the aforementioned ground feature information set to obtain a water dynamic information set. This allows for the avoidance of flooded areas in the access route. Finally, based on the aforementioned vegetation information set, farmland plot vector set, building risk point information set, and water dynamic information set, a transmission line vector layer set is generated. This provides a directly operable decision-making environment for subsequent route planning. Finally, based on the aforementioned transmission line vector map layer set, the access route for the transmission facilities is determined, and the aforementioned transportation equipment is controlled to complete the access operation of the transmission facilities according to the access route. Ultimately, this reduces the safety risks associated with the access of transmission facilities during the process.

[0085] Further reference Figure 2 As an implementation of the methods shown in the above figures, this disclosure provides some embodiments of a power transmission facility access device based on remote sensing image recognition. These device embodiments are similar to... Figure 2 Corresponding to the method embodiments shown, the device can be specifically applied to various electronic devices.

[0086] like Figure 2 As shown, some embodiments of the power transmission facility access device 200 based on remote sensing image recognition include: an acquisition unit 201, a fusion correction processing unit 202, a ground feature semantic segmentation unit 203, a vegetation point cloud segmentation processing unit 204, a crop type clustering unit 205, a first determination unit 206, a water area dynamic analysis unit 207, a generation unit 208, and a second determination unit 209. The acquisition unit 201 is configured to acquire a remote sensing image dataset of the target power transmission channel through the multi-source acquisition system in response to receiving a power transmission facility access command; the fusion correction processing unit 202 is configured to perform fusion correction processing on the remote sensing image dataset to obtain a fused geographic information dataset; the ground feature semantic segmentation unit 203 is configured to perform ground feature semantic segmentation on the fused geographic information dataset using a remote sensing image segmentation model to obtain a ground feature element information set; the vegetation point cloud segmentation processing unit 204 is configured to perform vegetation point cloud segmentation processing on the ground feature element information set to obtain a vegetation information set; and the crop type clustering unit 205 is configured to perform crop type clustering on the ground feature element information set to obtain farmland plots. Vector set; First determining unit 206, configured to determine the building outline information of the above-mentioned land feature information set, obtain the building outline set, and generate a building risk point information set based on the above-mentioned building outline set; Water dynamic analysis unit 207, configured to perform water dynamic analysis on the above-mentioned land feature information set, obtain a water dynamic information set; Generation unit 208, configured to generate a power transmission line vector layer set based on the above-mentioned vegetation information set, the above-mentioned farmland plot vector set, the above-mentioned building risk point information set, and the above-mentioned water dynamic information set; Second determining unit 209, configured to determine the power transmission facility access route based on the above-mentioned power transmission line vector layer set, and control the above-mentioned transportation equipment to complete the access operation of the above-mentioned power transmission facility based on the above-mentioned power transmission facility access route.

[0087] It is understandable that the units described in the device 200 are related to the reference. Figure 2 The steps in the described method correspond to each other. Therefore, the operations, features, and beneficial effects described above for the method also apply to the device 200 and the units contained therein, and will not be repeated here.

[0088] The following is for reference. Figure 3 It shows a schematic diagram of the structure of an electronic device (e.g., an electronic device) 300 suitable for implementing some embodiments of the present disclosure. Figure 3 The electronic device shown is merely an example and should not be construed as limiting the functionality and scope of the embodiments of this disclosure.

[0089] like Figure 3As shown, the electronic device 300 may include a processing unit (e.g., a central processing unit, a graphics processing unit, etc.) 301, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 302 or a program loaded from a storage device 308 into a random access memory (RAM) 303. The RAM 303 also stores various programs and data required for the operation of the electronic device 300. The processing unit 301, ROM 302, and RAM 303 are interconnected via a bus 304. An input / output (I / O) interface 305 is also connected to the bus 304.

[0090] Typically, the following devices can be connected to I / O interface 305: input devices 306 including, for example, touchscreens, touchpads, keyboards, mice, cameras, microphones, accelerometers, gyroscopes, etc.; output devices 307 including, for example, liquid crystal displays (LCDs), speakers, vibrators, etc.; storage devices 308 including, for example, magnetic tapes, hard disks, etc.; and communication devices 309. Communication device 309 allows electronic device 300 to communicate wirelessly or wiredly with other devices to exchange data. Although Figure 3 An electronic device 300 with various devices is shown; however, it should be understood that it is not required to implement or possess all of the devices shown. More or fewer devices may be implemented or possessed alternatively. Figure 3 Each box shown can represent a device or multiple devices as needed.

[0091] In particular, according to some embodiments of this disclosure, the processes described above with reference to the flowcharts can be implemented as computer software programs. For example, some embodiments of this disclosure include a computer program product comprising a computer program carried on a computer-readable medium, the computer program containing program code for performing the methods shown in the flowcharts. In such embodiments, the computer program can be downloaded and installed from a network via communication device 309, or installed from storage device 308, or installed from ROM 302. When the computer program is executed by processing device 301, it performs the functions defined in the methods of some embodiments of this disclosure.

[0092] It should be noted that, in some embodiments of this disclosure, the computer-readable medium described above may be a computer-readable signal medium or a computer-readable storage medium, or any combination thereof. A computer-readable storage medium may be, for example,—but not limited to—an electrical, magnetic, optical, electromagnetic, infrared, or semiconductor system, apparatus, or device, or any combination thereof. More specific examples of a computer-readable storage medium may include, but are not limited to: an electrical connection having one or more wires, a portable computer disk, a hard disk, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fiber, portable compact disk read-only memory (CD-ROM), optical storage device, magnetic storage device, or any suitable combination thereof. In some embodiments of this disclosure, a computer-readable storage medium may be any tangible medium containing or storing a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. In some embodiments of this disclosure, a computer-readable signal medium may include a data signal propagated in baseband or as part of a carrier wave, carrying computer-readable program code. Such propagated data signals may take various forms, including but not limited to electromagnetic signals, optical signals, or any suitable combination thereof. A computer-readable signal medium can be any computer-readable medium other than a computer-readable storage medium, which can send, propagate, or transmit a program for use by or in connection with an instruction execution system, apparatus, or device. The program code contained on the computer-readable medium can be transmitted using any suitable medium, including but not limited to: wires, optical fibers, RF (radio frequency), etc., or any suitable combination thereof.

[0093] In some implementations, clients and servers can communicate using any currently known or future-developed network protocol such as HTTP (Hypertext Transfer Protocol) and can interconnect with digital data communication (e.g., communication networks) of any form or medium. Examples of communication networks include local area networks (“LANs”), wide area networks (“WANs”), the Internet (e.g., the Internet of Things), and peer-to-peer networks (e.g., ad hoc peer-to-peer networks), as well as any currently known or future-developed networks.

[0094] The aforementioned computer-readable medium may be included in the aforementioned electronic device; or it may exist independently and not assembled into the electronic device. The aforementioned computer-readable medium carries one or more programs, which, when executed by the electronic device, cause the electronic device to: in response to receiving a power transmission facility approach command, acquire a remote sensing image dataset of the target power transmission channel through the aforementioned multi-source acquisition system; perform fusion correction processing on the aforementioned remote sensing image dataset to obtain a fused geographic information dataset; perform semantic segmentation of the aforementioned fused geographic information dataset using a remote sensing image segmentation model to obtain a set of geographic feature information; perform vegetation point cloud segmentation processing on the aforementioned geographic feature information set to obtain a vegetation information set; and perform agricultural processing on the aforementioned geographic feature information set. Clustering of land features yields a vector set of farmland plots; building outline information of the aforementioned land feature information set is determined to obtain a building outline set, and a building risk point information set is generated based on the aforementioned building outline set; dynamic analysis of water areas is performed on the aforementioned land feature information set to obtain a dynamic water area information set; a vector layer set of power transmission lines is generated based on the aforementioned vegetation information set, the aforementioned farmland plot vector set, the aforementioned building risk point information set, and the aforementioned dynamic water area information set; the access route of the power transmission facilities is determined based on the aforementioned access route of the power transmission facilities, and the aforementioned transportation equipment is controlled to complete the access operation of the aforementioned power transmission facilities based on the aforementioned access route of the power transmission facilities.

[0095] Computer program code for performing operations of some embodiments of this disclosure can be written in one or more programming languages ​​or a combination thereof, including object-oriented programming languages ​​such as Java, Smalltalk, and C++, and conventional procedural programming languages ​​such as the "C" language or similar programming languages. The program code can be executed entirely on the user's computer, partially on the user's computer, as a standalone software package, partially on the user's computer and partially on a remote computer, or entirely on a remote computer or server. In cases involving remote computers, the remote computer can be connected to the user's computer via any type of network—including a local area network (LAN) or a wide area network (WAN)—or can be connected to an external computer (e.g., via the Internet using an Internet service provider).

[0096] The flowcharts and block diagrams in the accompanying drawings illustrate the architecture, functionality, and operation of possible implementations of systems, methods, and computer program products according to various embodiments of this disclosure. In this regard, each block in a flowchart or block diagram may represent a module, segment, or portion of code containing one or more executable instructions for implementing a specified logical function. It should also be noted that in some alternative implementations, the functions indicated in the blocks may occur in a different order than those indicated in the drawings. For example, two consecutively indicated blocks may actually be executed substantially in parallel, and they may sometimes be executed in reverse order, depending on the functions involved. It should also be noted that each block in the block diagrams and / or flowcharts, and combinations of blocks in the block diagrams and / or flowcharts, can be implemented using a dedicated hardware-based system that performs the specified function or operation, or using a combination of dedicated hardware and computer instructions.

[0097] The units described in some embodiments of this disclosure can be implemented in software or hardware. The described units can also be housed in a processor; for example, a processor may be described as including an acquisition unit, a fusion correction processing unit, a ground feature semantic segmentation unit, a vegetation point cloud segmentation processing unit, a crop type clustering unit, a first determination unit, a water area dynamic analysis unit, a generation unit, and a second determination unit. The names of these units do not necessarily limit the specific unit; for example, the acquisition unit may also be described as "a unit that, in response to receiving a power transmission facility access command, acquires a remote sensing image dataset of the target power transmission channel through the aforementioned multi-source acquisition system."

[0098] The functions described above in this document can be performed at least in part by one or more hardware logic components. For example, exemplary types of hardware logic components that can be used, without limitation, include: field-programmable gate arrays (FPGAs), application-specific integrated circuits (ASICs), application-specific standard products (ASSPs), system-on-a-chip (SoCs), complex programmable logic devices (CPLDs), and so on.

[0099] The above description is merely a selection of preferred embodiments of this disclosure and an explanation of the technical principles employed. Those skilled in the art should understand that the scope of the invention involved in the embodiments of this disclosure is not limited to technical solutions formed by specific combinations of the above-described technical features, but should also cover other technical solutions formed by arbitrary combinations of the above-described technical features or their equivalents without departing from the above-described inventive concept. For example, technical solutions formed by substituting the above-described features with (but not limited to) technical features with similar functions disclosed in the embodiments of this disclosure.

Claims

1. A method for power transmission facility access based on remote sensing image recognition, applied to a power transmission facility access system, the power transmission facility access system comprising: A multi-source acquisition system and transportation equipment, the method comprising: In response to receiving a power transmission facility access command, the system acquires a remote sensing image dataset of the target power transmission channel through the multi-source acquisition system. The remote sensing image dataset is fused and corrected to obtain a fused geographic information dataset. Using a remote sensing image segmentation model, the fused geographic information dataset is subjected to semantic segmentation of ground features to obtain a set of ground feature information. The vegetation point cloud segmentation process is performed on the aforementioned ground feature information set to obtain a vegetation information set. The set of land feature information is clustered by crop type to obtain a vector set of farmland plots; The building outline information of the land feature information set is determined to obtain the building outline set, and a building risk point information set is generated based on the building outline set; A water area dynamics analysis is performed on the aforementioned set of land feature information to obtain a set of water area dynamics information. Based on the vegetation information set, the farmland plot vector set, the building risk point information set, and the water area dynamic information set, a power transmission line vector layer set is generated. Based on the vector map layer set of the power transmission line, the access route of the power transmission facility is determined, and based on the access route of the power transmission facility, the transportation equipment is controlled to complete the access operation of the power transmission facility.

2. The method according to claim 1, wherein, The remote sensing image dataset includes: visible light image data, thermal infrared image data, and point cloud data.

3. The method according to claim 1, wherein, The step of generating a transmission line vector layer set based on the vegetation information set, the farmland plot vector set, the building risk point information set, and the water area dynamic information set includes: The vegetation information set, the farmland plot vector set, the building risk point information set, and the water area dynamic information set are transformed by coordinate system to obtain a multi-source dataset. The multi-source dataset is used to perform three-dimensional modeling to obtain a three-dimensional power transmission scene model of the target power transmission channel; The three-dimensional power transmission scene model is encapsulated to obtain a three-dimensional geographic entity set; Spatial association is performed on the three-dimensional geographic entity set to obtain a vector layer set of power transmission lines.

4. The method according to claim 1, wherein, The step of determining the building outline information of the land feature information set to generate a building risk point information set includes: Extract the building pixel regions from the set of ground feature information to obtain the initial building outline vector set; The initial building outline vector set is regularized to obtain a regularized building outline set; Based on the aforementioned set of standardized building outlines, a set of building type labels is generated; Based on the building type tag set, a building risk point information set is generated.

5. The method according to claim 1, wherein, The process of fusing and correcting the remote sensing image dataset to obtain a fused geographic information dataset includes: Based on the aforementioned remote sensing image dataset, a multi-source image dataset is constructed; Geometric calibration and spatial registration are performed on the multi-source image dataset to obtain a registered image dataset; The registered image dataset is subjected to radiometric calibration and atmospheric correction to obtain a surface reflectance image set; The surface reflectance image set is subjected to topographic radiometric correction to obtain a topographically corrected image set. The terrain-corrected image set is subjected to cloud removal and shadow compensation processing to obtain a cloudless image set; The cloudless image set is subjected to image fusion processing to obtain a fused image set; Temporal filtering and feature enhancement processing are performed on the fused image set to obtain a fused geographic information dataset.

6. The method according to claim 1, wherein, The step of performing vegetation point cloud segmentation processing on the land feature information set to obtain a vegetation information set includes: Based on the aforementioned set of land feature information, the patch extent of vegetation is extracted to obtain a vector information set of vegetation area; Obtain point cloud data that matches the vector information set of the vegetation area to obtain the original point cloud dataset; The original point cloud dataset is normalized to obtain a normalized vegetation point cloud dataset; The normalized vegetation point cloud set is segmented into individual tree point clouds using a single tree segmentation algorithm to obtain individual tree point cloud clusters. Extract the geometric features and spectral reflectance features of each individual tree point cloud cluster in the individual tree point cloud cluster set to obtain the individual tree feature vector set; Based on the single tree feature vector set, generate a single tree species label set; A vegetation information set is generated based on the single tree point cloud cluster and the single tree species tag set.

7. A power transmission facility access device based on remote sensing image recognition, comprising: The acquisition unit is configured to acquire a remote sensing image dataset of the target power transmission channel through the multi-source acquisition system in response to receiving a power transmission facility access command. The fusion correction processing unit is configured to perform fusion correction processing on the remote sensing image dataset to obtain a fused geographic information dataset. The semantic segmentation unit is configured to perform semantic segmentation of the fused geographic information dataset using a remote sensing image segmentation model to obtain a set of geographic feature information. The vegetation point cloud segmentation processing unit is configured to perform vegetation point cloud segmentation processing on the land feature information set to obtain a vegetation information set. The crop type clustering unit is configured to perform crop type clustering on the set of land feature information to obtain a vector set of farmland plots; The first determining unit is configured to determine the building outline information of the land feature information set, obtain the building outline set, and generate a building risk point information set based on the building outline set. The water area dynamic analysis unit is configured to perform water area dynamic analysis on the land feature information set to obtain a water area dynamic information set. The generation unit is configured to generate a vector layer set of power transmission lines based on the vegetation information set, the farmland plot vector set, the building risk point information set, and the water dynamic information set. The second determining unit is configured to determine the access route of the power transmission facility based on the power transmission line vector map layer set, and control the transport equipment to complete the access operation of the power transmission facility based on the access route of the power transmission facility.

8. An electronic device, comprising: One or more processors; A storage device on which one or more programs are stored; When the one or more programs are executed by the one or more processors, the one or more processors implement the method as described in any one of claims 1 to 6.

9. A computer-readable medium having a computer program stored thereon, wherein, When the program is executed by the processor, it implements the method as described in any one of claims 1 to 6.

Images