Method and system for optimizing cadastral surveying and mapping based on low-altitude remote sensing

Abstract

The invention discloses a cadastral surveying and mapping optimization method and system based on low-altitude remote sensing, and relates to the technical field of surveying and mapping optimization, and the method comprises the steps: collecting the satellite image data of a target measurement area, and recognizing the building density distribution, the vegetation coverage density distribution and the terrain height difference change rate; carrying out unmanned aerial vehicle type and sensor type selection in a preset resource library, and then executing flight path planning based on the flight height, the course overlapping ratio and the lateral overlapping ratio; controlling the unmanned aerial vehicle to fly, collecting a low-altitude remote sensing data set and establishing a live-action three- And calling the cadastral mapping knowledge domain to perform reasoning, verification and conflict detection, and generating a target cadastral map. The technical problems of low cadastral element recognition accuracy and reliability and low cadastral surveying and mapping operation efficiency caused by difficulty in accurate data acquisition in a load environment in cadastral surveying and mapping operation in the prior art are solved. The technical effects of improving the cadastral surveying and mapping operation efficiency and enhancing the accuracy and reliability of cadastral element recognition in a complex environment are achieved.

Description

Technical Field

[0001] This invention relates to the field of surveying and mapping optimization technology, specifically to a cadastral surveying and mapping optimization method and system based on low-altitude remote sensing. Background Technology

[0002] Traditional cadastral surveying primarily relies on ground-based surveying methods such as total stations and GPS-RTK. This method has a limited operational range and is easily affected by factors such as terrain, vegetation, and building obstruction. Furthermore, data acquisition efficiency drops significantly in densely built-up areas, areas with dense vegetation, or areas with significant terrain undulations, making it difficult to meet the demands of modern cadastral management for high efficiency, real-time data, and comprehensive data. While UAVs can carry various sensors for low-altitude data acquisition, optimizing surveying paths based on specific characteristics of the survey area, such as building density, vegetation cover, and terrain undulations, remains a key challenge. Additionally, it is still difficult to automatically extract cadastral elements, such as boundary points, boundaries, and plot boundaries, from 3D models. Furthermore, issues such as blurred feature boundaries, model gaps due to obstruction, and insufficient semantic information may also affect the accuracy and reliability of cadastral surveying.

[0003] Therefore, in the current related technologies, there are technical problems such as the difficulty in accurately collecting data in cadastral surveying operations under heavy load conditions, resulting in low accuracy and reliability of cadastral element identification and low efficiency of cadastral surveying operations. Summary of the Invention

[0004] This application provides an optimized cadastral surveying method and system based on low-altitude remote sensing, which solves the technical problem in the prior art that cadastral surveying operations are difficult to accurately collect data under heavy load conditions, resulting in low accuracy and reliability of cadastral element identification and low efficiency of cadastral surveying operations. It achieves the technical effect of improving the efficiency of cadastral surveying operations and enhancing the accuracy and reliability of cadastral element identification in complex environments.

[0005] This application provides a cadastral mapping optimization method based on low-altitude remote sensing. The method includes: acquiring satellite image data of a target survey area, identifying the building density distribution, vegetation cover density distribution, and topographic elevation change rate of the target survey area, and generating an identification feature set; based on the identification feature set, selecting UAV type and sensor type from a pre-set resource library, and then performing flight path planning based on flight altitude, heading overlap rate, and lateral overlap rate based on the selection results to establish a UAV-sensor-path mapping; controlling the UAV flight with the UAV-sensor-path mapping, acquiring low-altitude remote sensing datasets to establish a real-scene 3D model of the target survey area; calling a cadastral knowledge graph to perform reasoning, verification, and conflict detection on the real-scene 3D model, completing the correction of cadastral elements, and generating a target cadastral map.

[0006] In a possible implementation, the cadastral mapping optimization method based on low-altitude remote sensing further performs the following processing: From the satellite image data, identify building outlines through semantic segmentation, calculate the number of buildings or coverage rate per unit area, and obtain the building density distribution; Based on the satellite image data, distinguish between vegetation and non-vegetation by calculating vegetation indices, and assess the density of vegetation to obtain the vegetation cover density distribution; Perform spatial analysis on the satellite image data, calculate the difference in elevation between each cell and surrounding cells, and generate a slope map reflecting the rate of change in terrain elevation; Generate the identification feature set using the building density distribution, the vegetation cover density distribution, and the slope map.

[0007] In a possible implementation, the cadastral mapping optimization method based on low-altitude remote sensing further performs the following processing: performing multi-level consistent clustering on the target survey area according to the identification feature set based on building density distribution, vegetation cover density distribution, and topographic elevation change rate to establish each sub-region; based on the pre-set resource library, performing mapping accuracy optimization under corresponding environmental conditions on each sub-region, completing the selection of UAV type and sensor type, and obtaining the selection results corresponding to each sub-region.

[0008] In a possible implementation, the cadastral mapping optimization method based on low-altitude remote sensing further performs the following processing: collecting a set of historical mapping records based on the preset resource library, where each historical mapping record corresponds to a set of resource combinations, and each set of resource combinations includes at least one UAV type and one sensor type; for each sub-region, matching each reference mapping record set whose regional environmental similarity reaches a preset similarity threshold in the historical mapping record set; and extracting the resource combination with the highest cadastral modeling accuracy from each reference mapping record set to obtain the various selection results.

[0009] In a possible implementation, the cadastral mapping optimization method based on low-altitude remote sensing further performs the following processing: 3D modeling complexity identification is performed based on the building density and vegetation cover density of each sub-region, generating various modeling complexity coefficients; a basic forward overlap rate threshold and a basic lateral overlap rate threshold are configured based on a preset flatness complexity threshold; enhancement weights for each sub-region are calculated using the modeling complexity coefficients based on the preset flatness complexity threshold, and the basic forward overlap rate threshold and the basic lateral overlap rate threshold are weighted to generate forward overlap rates and lateral overlap rates for each sub-region; flight altitude is configured based on the distance constraints of sensors within each selection result, and flight path planning is performed in conjunction with the forward overlap rate and lateral overlap rate of each sub-region to generate planned paths for each sub-region; a correspondence is constructed between the selection results and the planned paths for each sub-region to generate the UAV-sensor-path mapping.

[0010] In a possible implementation, the cadastral mapping optimization method based on low-altitude remote sensing further performs the following processing: the basic forward overlap rate threshold and the basic lateral overlap rate threshold are the lower limits of the forward overlap rate and the lower limits of the lateral overlap rate of the preset flat area when the three-dimensional modeling accuracy meets the preset accuracy threshold, wherein the three-dimensional modeling complexity of the preset flat area is less than the preset flatness complexity threshold.

[0011] In a possible implementation, the cadastral mapping optimization method based on low-altitude remote sensing further performs the following processing: generating enhancement weights for each sub-region by calculating the enhancement ratio of each modeling complexity coefficient relative to the preset flatness complexity threshold.

[0012] In a possible implementation, the cadastral mapping optimization method based on low-altitude remote sensing further performs the following processing: extracting cadastral elements from the real-world 3D model; comparing the extracted cadastral elements with predefined ownership delimitation rules in the cadastral knowledge graph to determine if there is any ambiguity; if so, marking the ambiguous areas in the real-world 3D model and generating prompt information; sending the prompt information to a human-computer interaction interface to a manual verification terminal and receiving the returned correction instructions; correcting the cadastral elements according to the correction instructions, and generating a target cadastral map from the corrected real-world 3D model.

[0013] In a possible implementation, the cadastral mapping optimization method based on low-altitude remote sensing also performs the following processing: the cadastral knowledge graph is obtained by identifying and structuring cadastral entities and the relationships between them, based on collected multi-source cadastral rule data.

[0014] This application also provides a cadastral mapping optimization system based on low-altitude remote sensing. The system includes: a feature recognition module, used to collect satellite image data of the target survey area, identify the building density distribution, vegetation cover density distribution, and topographic elevation change rate of the target survey area, and generate a recognition feature set; a path planning module, used to select UAV type and sensor type from a pre-set resource library based on the recognition feature set, and then perform flight path planning based on flight altitude, heading overlap rate, and lateral overlap rate based on the selection results, and establish a UAV-sensor-path mapping; a model building module, used to control the flight of the UAV with the UAV-sensor-path mapping, collect low-altitude remote sensing datasets to build a real-scene 3D model of the target survey area; and a cadastral map generation module, used to call a cadastral knowledge graph, perform reasoning, verification, and conflict detection on the real-scene 3D model, complete the correction of cadastral elements, and generate a target cadastral map.

[0015] This application proposes a low-altitude remote sensing-based cadastral mapping optimization method and system. This method involves collecting satellite imagery data of the target survey area to identify building density distribution, vegetation cover density distribution, and topographic elevation variation rate. The system then selects UAV and sensor types from a pre-set resource library and performs flight path planning based on flight altitude, forward overlap rate, and lateral overlap rate. It controls the UAV to collect low-altitude remote sensing datasets to build a realistic 3D model. Finally, it calls upon a cadastral knowledge graph for reasoning, verification, and conflict detection to generate the target cadastral map. This approach solves the technical problems in existing cadastral mapping operations where accurate data collection under heavy load conditions is difficult, leading to low accuracy and reliability of cadastral element identification and low efficiency in cadastral mapping operations. It achieves the technical effect of improving the efficiency of cadastral mapping operations and enhancing the accuracy and reliability of cadastral element identification in complex environments. Attached Figure Description

[0016] To more clearly illustrate the technical solutions of the embodiments of this disclosure, the accompanying drawings of the embodiments of this disclosure will be briefly described below. Flowcharts are used in this application to illustrate the operations performed by the system according to the embodiments of this application. It should be understood that the preceding or following operations are not necessarily performed precisely in sequence. Instead, various steps can be processed in reverse order or simultaneously as needed. Furthermore, other operations can be added to these processes, or one or more steps can be removed from these processes.

[0017] Figure 1 This is a schematic diagram of the cadastral mapping optimization method based on low-altitude remote sensing provided in an embodiment of this application.

[0018] Figure 2 This is a schematic diagram of the structure of the cadastral mapping optimization system based on low-altitude remote sensing provided in an embodiment of this application.

[0019] Figure labeling: Feature recognition module 10, path planning module 20, model building module 30, cadastral map generation module 40. Detailed Implementation

[0020] To further illustrate the technical means and effects of the present invention in achieving its intended purpose, the following detailed description of the specific implementation methods, structures, features, and effects of the present invention, in conjunction with the accompanying drawings and preferred embodiments, is provided below.

[0021] This application provides a cadastral mapping optimization method based on low-altitude remote sensing, such as... Figure 1 As shown, the method includes: Step S100: Collect satellite image data of the target survey area, identify the building density distribution, vegetation cover density distribution, and topographic elevation change rate of the target survey area, and generate an identification feature set.

[0022] Step S100 further includes step S110, identifying building outlines from the satellite image data through semantic segmentation, calculating the number of buildings or coverage rate per unit area, and obtaining the building density distribution; step S120, based on the satellite image data, distinguishing between vegetation and non-vegetation by calculating the vegetation index, and assessing the density of vegetation, and obtaining the vegetation cover density distribution; step S130, performing spatial analysis on the satellite image data, calculating the difference between the elevation of each cell and the surrounding cells, and generating a slope map reflecting the rate of change of terrain elevation; step S140, generating the identification feature set using the building density distribution, the vegetation cover density distribution, and the slope map.

[0023] Preferably, according to the requirements of the surveying and mapping mission, the target survey area is imaged by a remote sensing satellite equipped with an optical sensor to obtain satellite image data, including panchromatic and multispectral bands, for identifying fine contour features and attributes of ground objects. Then, the building density distribution, vegetation cover density distribution, and topographic elevation change rate of the target survey area are identified. Specifically, semantic segmentation is performed on the satellite image data, that is, pixel-level classification processing of the satellite image data is performed using a pre-trained convolutional neural network to identify building outlines, including labeling each pixel in the satellite image as a building or non-building, and then outputting a binary mask image, where white pixels represent buildings. Black pixels represent non-buildings. Then, fixed-size grid cells, such as 100m×100m, are defined on the generated building mask image. Within each grid cell, the number of building outlines is identified and counted to obtain the number of buildings per unit area. The ratio of the total number of pixels classified as buildings in each grid cell to the total number of pixels in that grid cell is calculated as the coverage rate. The number or coverage rate of each grid cell is then assigned to that cell to generate a building density distribution covering the entire target survey area, with each grid cell having a corresponding density value. High-density areas indicate severe occlusion and require oblique photography to obtain building facade information, while low-density areas can be satisfied with vertical photography.

[0024] Preferably, the vegetation index is calculated using the multispectral bands of satellite imagery through the normalized difference vegetation index formula. This formula calculates the ratio of the difference between the reflectance values ​​of the near-infrared band and the red band to the sum of the reflectance values ​​of the near-infrared band and the red band. Then, by setting a vegetation index threshold, the image is binarized to distinguish between vegetation and non-vegetation. The vegetation index of vegetated areas is close to 1, while the vegetation index of non-vegetated areas is very low, or even negative. The vegetation index value directly reflects the chlorophyll content or the density of vegetation; the higher the value, the denser the vegetation. Similarly, the average vegetation index value of all pixels in each grid cell is calculated, and the calculation result is assigned to each grid cell to generate a vegetation cover density distribution, which reflects both the presence and density of vegetation.

[0025] Preferably, spatial analysis is performed on satellite imagery data to calculate slope, specifically calculating the difference in elevation between each cell and its surrounding cells. This involves traversing each pixel in the digital elevation model (DEM) data and calculating the maximum rate of change between its elevation value and the elevation values ​​of its eight adjacent cells. Specifically, the slope value is calculated by combining the derivatives of the elevation changes at that point in the east-west and north-south directions. Finally, a slope map reflecting the rate of change in terrain elevation is output, where the value of each pixel represents the inclination of the ground at that location. A larger slope value indicates a more dramatic change in terrain elevation, while a smaller slope value indicates a flatter terrain. Finally, the building density distribution, vegetation cover density distribution, and slope map are spatially registered to ensure alignment of each grid cell, generating a recognition feature set—the registered raster layer set. For any geographical location within the target detection area, the building density value, vegetation cover density value, and slope value of that point can be queried simultaneously.

[0026] Step S200: Based on the identification feature set, select the UAV type and sensor type from the preset resource library, and then perform flight path planning based on flight altitude, heading overlap rate and lateral overlap rate based on the selection results to establish UAV-sensor-path mapping.

[0027] Step S200 further includes step S210, performing multi-level consistent clustering on the target survey area according to the identification feature set based on building density distribution, vegetation cover density distribution, and terrain elevation change rate to establish each sub-region; step S220, based on the preset resource library, performing mapping accuracy optimization under corresponding environmental conditions on each sub-region, completing the selection of UAV type and sensor type, and obtaining the selection results corresponding to each sub-region.

[0028] Preferably, unsupervised clustering algorithms such as K-Means or DBSCAN are used to perform multi-level consistent clustering on the target survey area based on building density distribution, vegetation cover density distribution, and topographic elevation change rate according to the identification feature set. That is, pixels that are close in distance in the feature space are grouped into the same class. Specifically, building density value, vegetation cover density value, and slope value are considered simultaneously to generate a cluster distribution, ensuring that the clustering results comprehensively reflect the combined influence of all environmental factors. Each pixel is assigned a category label, and all pixels with the same label and spatially connected together constitute a sub-region. Finally, various sub-regions are established, such as urban dense sub-region A, high building density, low vegetation density, and low slope; forest and hilly sub-region B, low building density, high vegetation density, and high slope; and plain and farmland sub-region C, low building density, medium vegetation density, and low slope.

[0029] Preferably, the pre-set resource library includes historical surveying records. Each record contains at least the building density, vegetation density, and slope of the historical task area, the type of UAV and sensor used, and the accuracy evaluation results of the final generated real-world 3D model, such as the mean square error and elevation mean square error. Based on the pre-set resource library, the surveying accuracy of each sub-area is optimized under the corresponding environmental conditions. That is, each sub-area is traversed, and based on average environmental characteristics such as average building density, average vegetation density, and average slope, a search and matching is performed in the historical surveying records. By calculating Euclidean distance or cosine similarity, all records with historical environmental characteristics highly similar to the current sub-area are identified. Among all matched historical records, the record with the highest accuracy is identified as the optimal historical record. The combination of UAV and sensor types contained in the optimal historical record is determined as the selection result for the current sub-area. Finally, the selection results corresponding to each sub-area are obtained, ensuring the allocation of appropriate hardware resources to achieve optimal overall efficiency while ensuring accuracy.

[0030] Furthermore, step S220 also includes step S221, collecting a set of historical surveying records based on the preset resource library, where each historical surveying record corresponds to a set of resource combinations, and each set of resource combinations includes at least one UAV type and one sensor type; step S222, for each sub-region, matching each reference surveying record set whose regional environmental similarity reaches a preset similarity threshold in the set of historical surveying records; step S223, extracting the resource combination with the highest cadastral modeling accuracy from each reference surveying record set to obtain each selection result.

[0031] Preferably, a set of historical surveying records from a pre-set resource library is obtained. Each historical surveying record corresponds to a set of resource combinations, used to specify the hardware resources used to execute the historical task, including at least one type of UAV and one type of sensor; it also includes average environmental characteristics such as average building density, average vegetation density, and average slope, used to record the environmental parameters of the area where the historical task is located; and it also includes cadastral modeling accuracy, used to measure the accuracy of the final real-scene 3D model generated by the task on cadastral elements such as boundary points and boundaries, such as plane error and elevation error. The lower the error value, the higher the accuracy.

[0032] Preferably, for each sub-region, the environmental feature vector is compared with the regional environmental feature vector of each record in the historical record database using Euclidean distance or cosine similarity. Specifically, a normalized multi-dimensional feature vector is constructed for each historical mapping task and the current sub-region, including building density, vegetation density, and terrain slope. The values ​​of each dimension are the result after normalization by the maximum and minimum values, ranging from [0, 1]. Preset feature weights of 0.5 for building density, 0.3 for vegetation density, and 0.2 for terrain slope are used, and these feature weights can be optimized and updated based on experience. The similarity between two feature vectors is calculated using the weighted cosine similarity formula. Based on experimental data and historical cadastral mapping task data, a preset similarity threshold of 0.85 is set and is adjustable. All historical records whose similarity calculation results reach or exceed the preset similarity threshold are filtered out. That is, when the feature similarity is greater than or equal to 0.85, the two regions are considered to be similar. Based on similar domain environments, corresponding reference mapping record sets are formed, containing all historical task records similar to the current sub-region environment. Then, each reference mapping record set is sorted in ascending order according to cadastral modeling accuracy. The unified evaluation standard for cadastral modeling accuracy adopts the accuracy indicators in the cadastral survey procedure, including horizontal mean square error and vertical mean square error. The horizontal mean square error is used to check the horizontal deviation between the upper boundary points of the model and the high-precision measured boundary points, and the root mean square error is calculated. The vertical mean square error is used to check the deviation between the model elevation and the measured elevation. The qualified threshold for modeling accuracy is set as horizontal mean square error less than or equal to 5cm and vertical mean square error less than or equal to 10cm, which can be adjusted according to the cadastral level. Then, the resource combination in the first historical record of each reference mapping record set sequence that meets the qualified threshold for modeling accuracy and has the smallest horizontal mean square error is extracted, including UAV type selection and sensor type selection. Finally, this is determined as the selection result for the current sub-region.

[0033] Furthermore, step S200 also includes step S230, identifying the 3D modeling complexity based on the building density and vegetation cover density of each sub-region, and generating various modeling complexity coefficients; step S240, configuring a basic forward overlap rate threshold and a basic lateral overlap rate threshold based on a preset flatness complexity threshold; step S250, calculating the enhancement weight of each sub-region based on the preset flatness complexity threshold and the various modeling complexity coefficients, performing weight enhancement processing on the basic forward overlap rate threshold and the basic lateral overlap rate threshold, and generating the forward overlap rate and lateral overlap rate of each sub-region; step S260, configuring the flight altitude based on the distance constraints of sensors within each selection result, and performing flight path planning based on the forward overlap rate and lateral overlap rate of each sub-region, generating the planned path for each sub-region; step S270, constructing a correspondence between each selection result and the planned path for each sub-region, and generating the UAV-sensor-path mapping.

[0034] Step S250 further includes generating the enhancement weights for each sub-region by calculating the enhancement ratio of each modeling complexity coefficient relative to the preset flat complexity threshold.

[0035] Preferably, flight path planning is performed based on flight altitude, heading overlap rate, and lateral overlap rate based on the selection results. Specifically, the complexity of 3D modeling is identified based on the building density and vegetation cover density of each sub-region. That is, by calculating the weighted sum of building density and vegetation cover density, it is integrated into a value representing the complexity, and each modeling complexity coefficient is output. The larger the value, the more complex the 3D modeling of the sub-region. The weight coefficient is set according to historical experience. The denser the buildings, the more model details there are, and the higher the modeling complexity coefficient. The denser the vegetation, the more irregular its surface and the more it sways with the wind, the higher the modeling complexity coefficient.

[0036] Preferably, a preset flatness complexity threshold is used to define flat and simple regions. A basic forward overlap rate threshold and a basic lateral overlap rate threshold are configured based on the preset flatness complexity threshold. These are the minimum forward overlap rate between adjacent photos and the minimum lateral overlap rate between adjacent flight paths required to successfully complete 3D modeling in flat and simple regions. For example, the basic forward overlap rate threshold is 70% and the basic lateral overlap rate threshold is 60%. Then, based on the preset flatness complexity threshold, the enhancement weight of each sub-region is calculated using each modeling complexity coefficient. That is, the enhancement ratio of the sub-region modeling complexity coefficient to the preset flatness complexity threshold is calculated as the enhancement weight. The weight of a flat sub-region is 1, and the weight of a more complex sub-region is greater than 1. The basic forward overlap rate threshold and the basic lateral overlap rate threshold are then weighted by multiplying them by the enhancement weight to obtain the forward overlap rate and the lateral overlap rate of each sub-region.

[0037] Preferably, the flight altitude is configured based on the distance constraints of the sensors within each selection result. The sensor distance constraints refer to the physical parameters of the selected sensors, primarily focal length and pixel size. Combined with preset ground resolution requirements, a fixed relative flight altitude is calculated using the formula: Flight Altitude = (Focal Length * Preset Ground Resolution) / Pixel Size. Based on the flight altitude, flight path planning is performed, considering the forward overlap rate and lateral overlap rate of each sub-region. Specifically, a standard trajectory planning algorithm is used to automatically generate a set of parallel flight paths based on the sub-region boundaries, the set flight altitude, and the overlap rate, ensuring complete coverage of the entire sub-region and meeting customized overlap rate requirements. This results in the output of planned paths for each sub-region. Finally, a correspondence is established between each selection result and the planned paths for each sub-region. This involves mapping and associating the UAV type, sensor type, the planned path for the corresponding sub-region, and the flight altitude to generate a UAV-sensor-path mapping. This mapping guides different types of UAVs equipped with different sensors to perform cadastral mapping tasks in different sub-regions according to different planned paths.

[0038] Furthermore, step S240 also includes the following: the basic forward overlap rate threshold and the basic lateral overlap rate threshold are the lower limit of the forward overlap rate and the lower limit of the lateral overlap rate of the preset flat area when the 3D modeling accuracy meets the preset accuracy threshold, wherein the 3D modeling complexity of the preset flat area is less than the preset flatness complexity threshold.

[0039] Preferably, the preset flat area is a predefined ideal area with an extremely simple environment. Its 3D modeling complexity is less than the preset flatness complexity threshold, i.e., the building density is extremely low, almost zero; the vegetation cover density is extremely low, such as bare soil or hardened ground; the terrain elevation change rate is extremely low, and the terrain is flat. UAV aerial surveys are conducted in the preset flat area to complete 3D modeling. The accuracy of the final generated 3D model, including the planar mean square error and elevation mean square error, meets the preset accuracy threshold, i.e., the standard values ​​required by cadastral surveying specifications, such as a planar accuracy better than 5 cm. Then, through experimental analysis, the lower limits of the forward overlap rate and the lateral overlap rate are determined when the 3D modeling accuracy of the preset flat area meets the preset accuracy threshold. These are used as the basic forward overlap rate threshold and the basic lateral overlap rate threshold. For example, it has been verified that on an absolutely flat and open site, using a specific camera, setting the basic forward overlap rate to 70% and the basic lateral overlap rate to 60% is the minimum condition for successfully generating a 3D model that meets the accuracy requirements.

[0040] Step S300: Control the UAV flight using the UAV-sensor-path mapping, collect low-altitude remote sensing datasets, and establish a real-world 3D model of the target survey area.

[0041] Preferably, UAV flight is controlled by a UAV-sensor-path mapping system. Specifically, for each mapped sub-region task, equipment scheduling and loading are performed. Operators are prompted to attach the corresponding sensors to the designated UAV. The ground station control center uploads the corresponding flight path plan, including waypoints, altitude, speed, and flight parameters such as forward overlap and lateral overlap, to the corresponding UAV. The UAV flies along the planned flight path and controls the sensors to perform exposure and shooting at designated locations according to the set forward overlap and lateral overlap parameters. This ensures that each sub-region uses the most suitable equipment and optimized parameters for data acquisition, thereby obtaining low-altitude remote sensing data. The low-altitude remote sensing dataset mainly includes a large number of aerial images with high overlap and precise position and attitude data, as well as corresponding flight trajectory, attitude records and other flight auxiliary data. The collected low-altitude remote sensing dataset is then imported into oblique photogrammetry 3D modeling software for 3D modeling. This includes identifying the same feature points in each photo and optimizing the calculation of sparse point clouds by combining precise position and attitude data. High-precision dense matching is then performed to generate high-density 3D point cloud data. The 3D point cloud data is then connected to construct a triangular mesh model that represents the geometry of the ground surface. Finally, the color information of the aerial images is mapped onto the triangular mesh model to generate a realistic 3D model of the target survey area.

[0042] Step S400: Call the cadastral knowledge graph to perform reasoning, verification and conflict detection on the real-world 3D model, complete the correction of cadastral elements, and generate the target cadastral map.

[0043] Step S400 further includes step S410, extracting cadastral elements from the real-world 3D model; step S420, comparing the extracted cadastral elements with the predefined ownership delimitation rules in the cadastral knowledge graph to determine if there is any ambiguity. If so, marking the ambiguous area in the real-world 3D model and generating a prompt message; step S430, sending the prompt message to the manual verification terminal through the human-computer interaction interface and receiving the returned correction instruction; step S440, correcting the cadastral elements according to the correction instruction, and generating the target cadastral map from the corrected real-world 3D model.

[0044] Preferably, a cadastral knowledge graph is invoked to perform reasoning, verification, and conflict detection on the real-world 3D model. The cadastral knowledge graph is a graph storing cadastral rules and relationships. For example, building boundaries should maintain a distance of no less than 0.5 meters from land parcel boundaries; the boundaries of adjacent land parcels must coincide without gaps or overlaps; and walls are typically considered land parcel boundaries. Specifically, cadastral elements are extracted from the real-world 3D model based on semantic segmentation and edge detection, including building outlines, land parcel boundaries, linear features such as walls and fences, and road edges. Then, the extracted cadastral elements are compared with predefined ownership definition rules in the cadastral knowledge graph. That is, the predefined ownership definition rules in the cadastral knowledge graph are invoked to check each extracted cadastral element. When a violation of the ownership definition rules is detected, it is determined to be ambiguous. The defined ownership delineation rules include, but are not limited to, land parcel closure, the relationship between buildings and land parcels, boundary point consistency, and walls and boundaries. Land parcel closure means that the land parcel polygon must be closed and not self-intersecting. The relationship between buildings and land parcels means that the outline of a building must be completely located within a certain land parcel and maintain a minimum distance D from the land parcel boundary, such as D=0.5 meters. Boundary point consistency means that the shared boundary line of adjacent land parcels must be defined by the same sequence of boundary points. Walls and boundaries mean that if a wall is continuous and undisputed, it can be regarded as a land parcel boundary. For example, if the outline of automatically extracted building M overlaps with the boundary line of land parcel N, the ambiguous area will be marked in the real-world 3D model, such as by highlighting, annotating, or generating error markers and prompts, such as "Warning: A building element at coordinate xx is found to conflict with the land parcel boundary, involving an error in ownership delineation."

[0045] Preferably, the 3D model with ambiguous labeled areas and detailed prompts are pushed to a human verification end for verification via a human-computer interaction interface. This includes combining domain expertise, historical archives, and field survey records to map vector elements such as building facades, land parcels, and boundary lines extracted from the real-world 3D model into entity instances in a knowledge graph. The instantiated data is then imported into a rule engine, which traverses the rule set in the knowledge graph and performs spatial operations and logical judgments such as buffer analysis and overlay analysis to detect rule violations. If a rule violation is detected regarding the relationship between buildings and land parcels, it is automatically applied to the relevant rules. The boundary between the house and the land parcel is highlighted in the 3D model, triggering annotation and prompts, and generating a report stating "The boundary distance between conflicting house IDH001 and land parcel IDP005 is 0.2 meters < 0.5 meters, suspected of encroachment." The issue is then manually assessed, and correction instructions are generated, which are then returned through a human-computer interaction interface. Finally, cadastral elements are corrected according to the correction instructions. For example, the boundary line of the house is moved according to the auditor's dragging to maintain a compliant distance from the land parcel boundary. Ultimately, a corrected real-world 3D model is generated, and the target cadastral map is output, thereby significantly improving the efficiency of cadastral surveying and updating.

[0046] Furthermore, step S400 also includes the fact that the cadastral knowledge graph is obtained by identifying and structurally representing cadastral entities and the relationships between them, based on collected multi-source cadastral rule data.

[0047] Preferably, based on the collected multi-source cadastral rule data, the cadastral entities and the relationships between them are identified and structurally expressed. This transforms scattered and unstructured cadastral rules, expert experience, domain standards, local conventions, etc., into a structured and semantic knowledge network that is understandable and reasonable by computers, thus obtaining a cadastral knowledge graph. This graph is capable of reasoning, verifying, detecting conflicts, and correcting cadastral elements in real-world 3D models. In this graph, land parcels, boundary points, boundary lines, buildings, walls, fences, roads, etc., in the cadastral domain are considered as cadastral entities, and the spatial logical relationships and constraint rules between entities are used as the connecting edges of the knowledge graph.

[0048] In the above text, refer to Figure 1 A cadastral mapping optimization method based on low-altitude remote sensing according to embodiments of the present invention is described in detail. Next, reference will be made to... Figure 2 A cadastral mapping optimization system based on low-altitude remote sensing according to an embodiment of the present invention is described.

[0049] The cadastral mapping optimization system based on low-altitude remote sensing according to embodiments of the present invention addresses the technical problems in the prior art where accurate data acquisition is difficult under heavy load conditions, leading to low accuracy and reliability of cadastral feature identification and low efficiency of cadastral mapping operations. It achieves the technical effect of improving the efficiency of cadastral mapping operations and enhancing the accuracy and reliability of cadastral feature identification in complex environments. Figure 2 As shown, the cadastral mapping optimization system based on low-altitude remote sensing includes: feature recognition module 10, path planning module 20, model building module 30, and cadastral map generation module 40.

[0050] The feature recognition module 10 is used to collect satellite image data of the target survey area, identify the building density distribution, vegetation cover density distribution, and topographic elevation change rate of the target survey area, and generate a recognition feature set; the path planning module 20 is used to select UAV type and sensor type from a preset resource library based on the recognition feature set, and then perform flight path planning based on flight altitude, heading overlap rate, and lateral overlap rate based on the selection results, and establish a UAV-sensor-path mapping; the model building module 30 is used to control the flight of the UAV with the UAV-sensor-path mapping, collect low-altitude remote sensing datasets to build a real-scene 3D model of the target survey area; the cadastral map generation module 40 is used to call the cadastral knowledge graph, perform reasoning, verification and conflict detection on the real-scene 3D model, complete the correction of cadastral elements, and generate a target cadastral map.

[0051] The specific configuration of the feature recognition module 10 will be described in detail below. The feature recognition module 10 further includes: identifying building outlines from the satellite image data through semantic segmentation, calculating the number of buildings or coverage per unit area to obtain a building density distribution; distinguishing between vegetation and non-vegetation by calculating a vegetation index based on the satellite image data, and assessing the density of vegetation to obtain a vegetation cover density distribution; performing spatial analysis on the satellite image data, calculating the difference in elevation between each cell and surrounding cells, and generating a slope map reflecting the rate of change in terrain elevation; and generating the recognition feature set using the building density distribution, the vegetation cover density distribution, and the slope map.

[0052] The specific configuration of the path planning module 20 will be described in detail below. The path planning module 20 further includes: performing multi-level consistent clustering on the target survey area according to the identified feature set based on building density distribution, vegetation cover density distribution, and terrain elevation change rate to establish each sub-region; and performing mapping accuracy optimization under corresponding environmental conditions on each sub-region based on the preset resource library to complete the selection of UAV type and sensor type, and obtain the selection results corresponding to each sub-region.

[0053] The specific configuration of the path planning module 20 will be described in detail below. The path planning module 20 further includes: collecting a set of historical surveying records based on the pre-set resource library, where each historical surveying record corresponds to a set of resource combinations, and each set of resource combinations includes at least one UAV type and one sensor type; for each sub-region, matching each reference surveying record set whose regional environmental similarity reaches a preset similarity threshold with the historical surveying record set; and extracting the resource combination with the highest cadastral modeling accuracy from each reference surveying record set to obtain the various selection results.

[0054] The specific configuration of the path planning module 20 will be described in detail below. The path planning module 20 further includes: identifying the complexity of 3D modeling based on the building density and vegetation cover density of each sub-region, generating various modeling complexity coefficients; configuring a basic forward overlap rate threshold and a basic lateral overlap rate threshold based on a preset flatness complexity threshold; calculating enhancement weights for each sub-region based on the preset flatness complexity threshold and the various modeling complexity coefficients, performing weight enhancement processing on the basic forward overlap rate threshold and the basic lateral overlap rate threshold, generating forward overlap rates and lateral overlap rates for each sub-region; configuring flight altitude based on the distance constraints of sensors within each selection result, combining the forward overlap rate and lateral overlap rate of each sub-region to plan flight paths, generating planned paths for each sub-region; and constructing a correspondence between the various selection results and the planned paths for each sub-region to generate the UAV-sensor-path mapping.

[0055] The specific configuration of the path planning module 20 will be described in detail below. The path planning module 20 further includes: the basic heading overlap rate threshold and the basic lateral overlap rate threshold are the lower limit of heading overlap rate and the lower limit of lateral overlap rate of the preset flat area when the 3D modeling accuracy meets the preset accuracy threshold, wherein the 3D modeling complexity of the preset flat area is less than the preset flatness complexity threshold.

[0056] The specific configuration of the path planning module 20 will be described in detail below. The path planning module 20 further includes: generating enhancement weights for each sub-region by calculating the enhancement ratio of each modeling complexity coefficient relative to the preset flat complexity threshold.

[0057] The specific configuration of the cadastral map generation module 40 will be described in detail below. The cadastral map generation module 40 further includes: extracting cadastral elements from the real-world 3D model; comparing the extracted cadastral elements with predefined ownership delimitation rules in the cadastral knowledge graph to determine if there is any ambiguity; if so, marking the ambiguous area in the real-world 3D model and generating a prompt message; sending the prompt message to a human verification terminal through a human-computer interaction interface and receiving a returned correction instruction; correcting the cadastral elements according to the correction instruction, and generating a target cadastral map from the corrected real-world 3D model.

[0058] The specific configuration of the cadastral map generation module 40 will be described in detail below. The cadastral map generation module 40 further includes: the cadastral knowledge graph is obtained by identifying and structurally representing cadastral entities and the relationships between them based on collected multi-source cadastral rule data.

[0059] The cadastral mapping optimization system based on low-altitude remote sensing provided in this invention can execute the cadastral mapping optimization method based on low-altitude remote sensing provided in any embodiment of this invention, and has the corresponding functional modules and beneficial effects of the method.

[0060] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-disclosed technical content to create equivalent embodiments without departing from the scope of the present invention. Any modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A cadastral mapping optimization method based on low-altitude remote sensing, characterized in that, include: Collect satellite imagery data of the target survey area, identify the building density distribution, vegetation cover density distribution, and topographic elevation change rate of the target survey area, and generate an identification feature set; Based on the identified feature set, UAV type and sensor type are selected from a pre-set resource library. Then, based on the selection results, flight path planning is performed based on flight altitude, heading overlap rate, and lateral overlap rate to establish a UAV-sensor-path mapping, specifically including: According to the identification feature set, perform multi-level consistent clustering on the target survey area based on building density distribution, vegetation cover density distribution, and topographic elevation change rate to establish each sub-region; Based on the pre-set resource library, the mapping accuracy optimization under the corresponding environmental conditions is performed on each sub-region, the selection of UAV type and sensor type is completed, and the selection results corresponding to each sub-region are obtained. The complexity of 3D modeling is identified based on the building density and vegetation cover density of each sub-region, and various modeling complexity coefficients are generated. Configure the basic heading overlap rate threshold and basic lateral overlap rate threshold based on the preset flatness complexity threshold; Based on the preset flatness complexity threshold, the enhancement weights of each sub-region are calculated using the various modeling complexity coefficients. The basic forward overlap rate threshold and the basic lateral overlap rate threshold are then weighted to generate the forward overlap rate and lateral overlap rate of each sub-region. Flight altitude is configured based on the distance constraints of sensors within each selection result. Flight path planning is performed by combining the forward overlap rate and the lateral overlap rate of each sub-region, and the planned path for each sub-region is generated. The correspondence between the various selection results and the planned paths of each sub-region is constructed to generate the UAV-sensor-path mapping; The drone-sensor-path mapping is used to control the drone flight and collect low-altitude remote sensing datasets to build a real-world 3D model of the target survey area. The cadastral knowledge graph is invoked to perform reasoning, verification, and conflict detection on the real-world 3D model, thereby correcting the cadastral elements and generating the target cadastral map.

2. The cadastral mapping optimization method based on low-altitude remote sensing as described in claim 1, characterized in that, Satellite imagery data of the target survey area is collected, and the building density, vegetation cover density, and topographic elevation change rate of the target survey area are identified to generate an identification feature set, including: From the satellite imagery data, the building outlines are identified through semantic segmentation, and the number of buildings or coverage rate per unit area is calculated to obtain the building density distribution. Based on the satellite imagery data, vegetation and non-vegetation are distinguished by calculating vegetation indices, and the density of vegetation is assessed to obtain the vegetation cover density distribution. Spatial analysis is performed on the satellite image data to calculate the difference in elevation between each cell and its surrounding cells, generating a slope map that reflects the rate of change in terrain elevation. The identification feature set is generated using the building density distribution, the vegetation cover density distribution, and the slope map.

3. The cadastral mapping optimization method based on low-altitude remote sensing as described in claim 1, characterized in that, The selection results corresponding to each of the sub-regions are obtained, including: Collect a set of historical mapping records based on the pre-set resource library. Each historical mapping record corresponds to a set of resource combinations. Each set of resource combinations includes at least one UAV type and one sensor type. For each sub-region, the reference mapping record sets that match the regional environment similarity in the historical mapping record set and reach a preset similarity threshold; The resource combination with the highest cadastral modeling accuracy is extracted from each of the reference mapping record sets to obtain the various selection results.

4. The cadastral mapping optimization method based on low-altitude remote sensing as described in claim 1, characterized in that, Configure the basic heading overlap rate threshold and the basic lateral overlap rate threshold, including: The basic forward overlap rate threshold and the basic lateral overlap rate threshold are the lower limits of forward overlap rate and lateral overlap rate of the preset flat area when the 3D modeling accuracy meets the preset accuracy threshold, wherein the 3D modeling complexity of the preset flat area is less than the preset flatness complexity threshold.

5. The cadastral mapping optimization method based on low-altitude remote sensing as described in claim 4, characterized in that, Based on the preset flatness complexity threshold, the enhancement weights for each sub-region are calculated using the various modeling complexity coefficients, including: The enhancement weights for each sub-region are generated by calculating the enhancement ratio of each modeling complexity coefficient relative to the preset flat complexity threshold.

6. The cadastral mapping optimization method based on low-altitude remote sensing as described in claim 1, characterized in that, The cadastral knowledge graph is invoked to perform reasoning, verification, and conflict detection on the real-world 3D model, thereby correcting cadastral elements and generating a target cadastral map, including: Extract cadastral elements from the real-world 3D model; The extracted cadastral elements are compared with the predefined ownership definition rules in the cadastral knowledge graph to determine whether there is any ambiguity. If so, the ambiguous areas are marked in the real-world 3D model and prompt information is generated. The prompt message is sent to the manual verification terminal through the human-computer interaction interface and the correction instruction is received in return; The cadastral elements are corrected according to the correction instructions, and the target cadastral map is generated from the corrected real-world 3D model.

7. The cadastral mapping optimization method based on low-altitude remote sensing as described in claim 6, characterized in that, The cadastral knowledge graph is obtained by identifying and structuring cadastral entities and the relationships between them, based on collected multi-source cadastral rule data.

8. A cadastral mapping optimization system based on low-altitude remote sensing, characterized in that, The system is used to implement the cadastral mapping optimization method based on low-altitude remote sensing as described in any one of claims 1 to 7, and the system comprises: The feature recognition module is used to collect satellite image data of the target survey area, identify the building density distribution, vegetation cover density distribution, and topographic elevation change rate of the target survey area, and generate a recognition feature set. The path planning module is used to select UAV type and sensor type from a preset resource library based on the identification feature set, and then perform flight path planning based on flight altitude, heading overlap rate and lateral overlap rate based on the selection result to establish UAV-sensor-path mapping. The model building module is used to control the flight of the UAV using the UAV-sensor-path mapping and to collect low-altitude remote sensing datasets to build a real-scene 3D model of the target survey area. The cadastral map generation module is used to call the cadastral knowledge graph to perform reasoning, verification and conflict detection on the real-world 3D model, complete the correction of cadastral elements, and generate the target cadastral map.

Images