Airborne laser scanning real-time mapping method for three-dimensional live-action map in complex geographical environment assisted by single Beidou system

Abstract

The invention provides a single-Beidou-system-assisted airborne laser scanning real-time mapping method for a three-dimensional live-action map in a complex geographical environment, and the method depends on the high-precision positioning and time service capability of a single Beidou system, and combines the point cloud collection advantage of airborne laser scanning equipment. Aiming at the problems of large positioning deviation, high data redundancy and insufficient real-time performance of a traditional mapping technology in complex geographical environments (such as mountainous areas and canyons, forest coverage areas and urban dense building groups), the invention provides a mapping method for the complex geographical environments. An integrated technical process of Beidou data acquisition and preprocessing, laser point cloud acquisition and denoising, feature point extraction and matching, three-dimensional model construction and precision check and optimization is constructed, and efficient and high-precision real-time mapping of a three-dimensional live-action map in a complex environment is realized. The method is suitable for geological exploration, emergency rescue and urban planning scenes, effectively improves the efficiency and precision of three-dimensional mapping in a complex geographical environment, and solves the bottleneck problems that the traditional technology depends on a multi-satellite system, is greatly interfered by the environment and cannot output results in real time.

Description

Technical Field

[0001] This invention belongs to the fields of geographic information engineering and remote sensing mapping technology. Specifically, it relates to a method for real-time airborne laser scanning mapping of complex geographic environments assisted by a single Beidou system. It is applicable to the construction of 3D real-scene maps in complex geographic environments (mountainous canyons, forest-covered areas, and densely built-up urban areas) and can be applied to scenarios such as geological exploration, emergency rescue, urban planning, and transportation construction. Background Technology

[0002] 3D reality maps, as an important form of geographic information representation, play a crucial role in many fields. With technological advancements, airborne laser scanning technology, due to its advantages of rapid and large-scale surface data acquisition, has become one of the core methods for 3D mapping. However, in complex geographical environments, traditional airborne laser scanning mapping technology still faces many bottlenecks: Positioning relies on multiple satellite systems and has weak anti-interference capabilities: Traditional technologies mostly rely on dual-mode positioning of GPS and BeiDou. In mountainous canyons and densely built-up urban areas, GPS signals are easily blocked, leading to a decrease in positioning accuracy. If only a single GPS system is relied upon, it cannot meet the positioning needs in complex environments, and the independent application of a single BeiDou system has not yet formed a mature technical solution.

[0003] Point cloud data has high redundancy and low processing efficiency: In complex geographical environments, laser scanning easily collects a large number of interfering point clouds such as vegetation and shadows. Traditional denoising algorithms use fixed thresholds, which cannot adapt to the interference features of different terrains, resulting in a low effective data extraction rate and a significant decrease in subsequent modeling efficiency.

[0004] Insufficient real-time capability, unable to meet emergency needs: Traditional mapping processes require data preprocessing, modeling, and accuracy verification to be completed on the ground. The time from data collection to output is relatively long (usually several hours to several days), which is difficult to meet the real-time needs of emergency rescue, disaster monitoring, and other scenarios.

[0005] Coordinate transformation and accuracy control are difficult: Under complex terrain, the attitude of the airborne platform fluctuates greatly, resulting in a large deviation between the point cloud data and the geographic coordinates; traditional coordinate transformation uses fixed parameters, which cannot dynamically adapt to the terrain features of different areas, further reducing the accuracy of the mapping.

[0006] Currently, my country's BeiDou Navigation Satellite System has achieved global networking, possessing high-precision positioning and timing capabilities, providing a technological foundation for single-satellite system-assisted mapping. Therefore, it is necessary to construct an airborne laser scanning real-time mapping method based on a single BeiDou system to overcome the bottlenecks of traditional technologies in complex geographical environments and achieve efficient and high-precision 3D real-scene map output. Summary of the Invention

[0007] The purpose of this invention is to provide a method for real-time airborne laser scanning mapping of 3D real-scene maps in complex geographical environments assisted by a single BeiDou system. Relying on the high-precision positioning and timing capabilities of a single BeiDou system, combined with airborne laser scanning technology, and through innovative algorithm and process design, this method solves the problems of large positioning deviations, high data redundancy, and insufficient real-time performance of traditional mapping techniques in complex geographical environments. It achieves efficient and high-precision real-time mapping of 3D real-scene maps, meeting the needs of scenarios such as geological exploration and emergency rescue.

[0008] The technical solution of this invention is: a method for real-time mapping of 3D real-scene maps of complex geographical environments using airborne laser scanning, assisted by a single BeiDou system. This method acquires real-time positioning (longitude, latitude, and elevation) and timing data from an airborne platform using a single BeiDou system, combines this with surface point cloud data collected by an airborne laser scanner, and applies adaptive point cloud denoising algorithms, terrain feature constraint matching models, and dynamic coordinate transformation mechanisms to construct a real-time mapping technology system for 3D real-scene maps of complex geographical environments. The basic steps include: BeiDou data acquisition and preprocessing, laser point cloud acquisition and denoising, feature point extraction and matching, 3D model construction, and accuracy verification and optimization. Based on the differences in terrain features of different complex environments (mountains, forests, cities), the algorithm parameters are dynamically adjusted to achieve high-precision real-time mapping output of 3D real-scene maps.

[0009] The specific implementation steps of the above method are as follows: Step 1. Airborne equipment deployment and BeiDou system initialization: Install a single BeiDou receiver, laser scanner, IMU and data storage module on the airborne platform (UAV, fixed-wing aircraft), complete the time synchronization and parameter calibration of the BeiDou system and laser equipment, and set the BeiDou positioning sampling frequency and laser scanning resolution; Step 2. BeiDou positioning data preprocessing: Receive raw positioning data (including longitude, latitude, elevation and timestamp) from a single BeiDou system, eliminate ionospheric and tropospheric delay errors through Kalman filtering algorithm, and combine IMU data to compensate for the attitude (roll, pitch, yaw) fluctuations of the airborne platform to obtain high-precision real-time position and attitude parameters of the airborne platform; Step 3. Laser point cloud acquisition and adaptive denoising: The airborne laser scanner acquires surface point cloud data according to preset parameters. The data includes the three-dimensional coordinates of the points (relative to the airborne platform), grayscale values, and reflection intensity. Based on the spatial density and grayscale characteristics of the point cloud, an adaptive threshold denoising algorithm is applied to remove points that are obscured by vegetation or shaded by buildings, while retaining effective shape and feature point clouds. Step 4. Feature point extraction and terrain matching: The SIFT algorithm is used to extract terrain feature points (such as mountain tops, valleys, and cliff edges) and ground feature points (such as building corners and road edges) from the denoised point cloud. The number of feature points is no less than 5% of the total number of points. Through the terrain feature constraint matching model, the feature points are associated with the coordinate system of Beidou positioning to establish the mapping relationship between point cloud data and geographic coordinates. Step 5. Dynamic coordinate transformation and 3D modeling: Based on the WGS84 coordinate system of Beidou positioning, the point cloud data is converted into the target coordinate system (such as the National Geodetic Coordinate System 2000) through a dynamic coordinate transformation mechanism. The irregular triangular mesh modeling method is applied to construct a 3D terrain model, and the texture information of the ground features is superimposed to generate a 3D real-scene map. Step 6. Accuracy Check and Model Optimization: Select ground control points with known coordinates (no fewer than 3), calculate the coordinate deviation of the corresponding points in the 3D real-world map, and apply the least squares principle to correct the model parameters; if the deviation exceeds the preset threshold, return to Step 2 to re-optimize the BeiDou data preprocessing parameters until the accuracy requirements are met.

[0010] In Step 1, the deployment of airborne equipment must meet the conditions for receiving BeiDou signals. In forest-covered areas or densely built-up urban areas, the antenna height and angle of the BeiDou receiver need to be adjusted to ensure that at least 4 satellites are visible. The scanning resolution of the laser scanner is dynamically set according to the mapping accuracy requirements. A resolution of 0.1-0.5m is used in geological exploration scenarios, and a resolution of 0.5-1m is used in emergency rescue scenarios to improve acquisition efficiency.

[0011] In Step 2, error compensation in BeiDou positioning data preprocessing includes: correcting satellite orbit errors using BeiDou system ephemeris data, and applying an atmospheric delay model to eliminate the influence of the ionosphere and troposphere on the signal; the fusion of IMU data and BeiDou positioning data adopts a 20 Hz IMU loose coupling method, with a sampling frequency 2-5 times that of BeiDou positioning frequency, to compensate for the instantaneous attitude changes of the airborne platform in real time, obtaining a POS with a horizontal plane ≤0.3 m and an elevation ≤0.2 m.

[0012] In Step 3, the threshold setting of the adaptive point cloud denoising algorithm is based on the local density of the point cloud: for complex terrain areas such as mountainous canyons, a lower threshold is used to retain detailed points; for forest-covered areas, a higher threshold is used to remove vegetation interference points; the denoised data must meet the requirement of an effective point cloud retention rate ≥ 90%; for mountainous canyon areas, a threshold of 0.2 times the average density is used to remove isolated interference points; for forest-covered areas, a threshold of 0.6 times the average density is used to remove vegetation points; for urban building clusters, a threshold of 0.5 times the average density is used to remove shadow interference points; if the requirements are not met, the denoising threshold is readjusted until the requirements are met.

[0013] In Step 4, the terrain feature constraint matching model is constructed in the following way: using the airborne platform position determined by BeiDou positioning as a reference, combined with the scanning angle and distance of the laser scanner, Calculate the absolute geographic coordinates (WGS84 coordinate system) of the feature points to establish the spatial geometric relationship of the point cloud; by constraining the elevation difference and slope terrain parameters of the feature points, eliminate mismatched points caused by scanning angle deviation, and the matching accuracy needs to reach more than 95%.

[0014] In Step 5, the dynamic coordinate transformation mechanism is used: based on preset coordinate transformation parameters (such as seven parameters or four parameters), the transformation coefficient is dynamically adjusted in combination with the real-time position of Beidou positioning to ensure that point cloud data from different areas are seamlessly stitched together in the target coordinate system; during the 3D modeling process, for dense urban building clusters, building outline texture information needs to be superimposed to improve the visual recognition of the real scene map.

[0015] In Step 6, the selection of ground control points must cover the entire mapping area and be evenly distributed. The accuracy check uses the mean square error of the plane position and the mean square error of the elevation as evaluation indicators. If the mean square error of the plane is >0.5m or the mean square error of the elevation is >0.3m, the Kalman filter parameters in Step 2 or the feature point matching threshold in Step 4 need to be re-optimized until the accuracy meets the standard.

[0016] The present invention has the following advantages: 1. The method of this invention eliminates the dependence on multiple satellite systems. Relying on the global positioning capability of BeiDou-3, it can maintain high-precision positioning in complex geographical environments through antenna adjustment and error compensation, thus solving the problem of signal blockage in traditional technologies.

[0017] 2. The adaptive point cloud denoising algorithm used in the method of this invention dynamically adjusts the threshold according to terrain features, which greatly reduces data redundancy and improves modeling efficiency.

[0018] 3. The method of this invention constructs an integrated real-time process, which takes only a few hours from data acquisition to output of a 3D real-scene map. Compared with the traditional ground processing process, the real-time performance is significantly improved, and it can be applied to scenarios such as emergency rescue and disaster monitoring.

[0019] 4. The method of this invention uses a dynamic coordinate transformation mechanism to adapt to the coordinate requirements of different regions, ensuring seamless point cloud stitching; through ground control point verification and parameter optimization, the mapping accuracy is stable and meets the application requirements of multiple scenarios.

[0020] 5. The method of the present invention is applicable to complex geographical environments such as mountain valleys, forest-covered areas, and urban building complexes. It can also adjust the equipment parameters and algorithm thresholds according to the needs of the scenario, and be extended to fields such as marine surveying and mapping and glacier monitoring, and has strong versatility and flexibility. Attached Figure Description

[0021] Figure 1 is a flowchart of the method for real-time airborne laser scanning mapping of complex geographical environments using a single BeiDou system-assisted three-dimensional real-scene map according to the present invention. Figure 2 for Figure 1 A flowchart illustrating the process of constructing a 3D real-scene map of a city in a certain region; Figure 3 for Figure 1 A flowchart illustrating the construction process of a 3D real-scene map of a forest coverage area in a certain forest region; Figure 4 for Figure 1 A flowchart illustrating the process of constructing a 3D real-scene map of a mountain canyon. Specific implementation methods The following description, in conjunction with the accompanying drawings, uses "3D real-scene map construction" as an example to illustrate the implementation scheme of the present invention, but should not be construed as a limitation on the technical solution.

[0022] Example 1: As Figure 1 , Figure 2 As shown, construct a 3D real-scene map of a certain region's city by following these steps. Step 1. Airborne Equipment Deployment and BeiDou System Initialization: Install a single BeiDou receiver, laser scanner, IMU, and data storage module on the airborne platform UAV. Complete time synchronization and parameter calibration between the BeiDou system and the laser equipment, and set the BeiDou positioning sampling frequency and laser scanning resolution. The airborne equipment deployment must meet the BeiDou signal reception requirements. In densely built-up urban areas, the antenna height and angle of the BeiDou receiver need to be adjusted to ensure that at least four satellites are visible. The scanning resolution of the laser scanner is dynamically set according to the mapping accuracy requirements; a resolution of 0.1-0.5m is used for geological exploration scenarios. Step 2. BeiDou Positioning Data Preprocessing: Receive raw positioning data from a single BeiDou system (including longitude, latitude, elevation, and timestamp). Use a Kalman filter algorithm to eliminate ionospheric and tropospheric delay errors. Combine this with IMU data to compensate for attitude (roll, pitch, yaw) fluctuations of the airborne platform, obtaining high-precision real-time position and attitude parameters. Error compensation in BeiDou positioning data preprocessing includes: correcting satellite orbit errors using BeiDou ephemeris data; applying an atmospheric delay model to eliminate the influence of the ionosphere and troposphere on the signal; and fusing IMU data with BeiDou positioning data using a 20Hz loosely coupled IMU with a sampling frequency 2-5 times the BeiDou positioning frequency to compensate for instantaneous attitude changes of the airborne platform in real time, obtaining a POS with a horizontal plane ≤0.3 m and an elevation ≤0.2 m. Step 3. Laser Point Cloud Acquisition and Adaptive Denoising: The airborne laser scanner acquires surface point cloud data according to preset parameters. The data includes the three-dimensional coordinates, grayscale values, and reflection intensity of the points. Based on the spatial density and grayscale characteristics of the point cloud, an adaptive threshold denoising algorithm is applied to remove points obscured by vegetation and building shadows, retaining effective topographic and feature point clouds. The threshold setting of the adaptive point cloud denoising algorithm is based on the local density of the point cloud, and the denoised data must meet the requirement of an effective point cloud retention rate ≥90%. For urban building clusters, 0.5 times the average density is used as the threshold to remove shadow interference points. If the standard is not met, the denoising threshold is readjusted until the requirements are met. Step 4. Feature Point Extraction and Terrain Matching: The SIFT algorithm is used to extract terrain feature points and ground feature points (such as building corners and road edges) from the denoised point cloud. The number of feature points should not be less than 5% of the total number of points. Through a terrain feature constraint matching model, the feature points are associated with the coordinate system of BeiDou positioning to establish a mapping relationship between point cloud data and geographic coordinates. The terrain feature constraint matching model is constructed as follows: using the location of the airborne platform of BeiDou positioning as the reference, combined with the scanning angle and distance of the laser scanner, the absolute geographic coordinates (WGS84 coordinate system) of the feature points are calculated to establish the spatial geometric relationship of the point cloud. Through the elevation difference and slope terrain parameters of the feature points, mismatched points caused by scanning angle deviation are eliminated, and the matching accuracy should reach more than 95%. Step 5. Dynamic Coordinate Transformation and 3D Modeling: Based on the WGS84 coordinate system of BeiDou positioning, point cloud data is converted into a target coordinate system (such as the National Geodetic Coordinate System 2000) through a dynamic coordinate transformation mechanism. An irregular triangular mesh modeling method is applied to construct a 3D terrain model, and texture information of ground features is superimposed to generate a 3D real-world map. The dynamic coordinate transformation mechanism is based on preset coordinate transformation parameters (such as seven parameters) and dynamically adjusts the transformation coefficients according to the real-time position of BeiDou positioning to ensure seamless stitching of point cloud data from different areas in the target coordinate system. During the 3D modeling process, for dense urban building clusters, building outline texture information needs to be superimposed to improve the visual recognizability of the real-world map. Step 6. Accuracy Check and Model Optimization: Select ground control points with known coordinates (at least 3), calculate the coordinate deviation of the corresponding points in the 3D real-world map, and apply the least squares principle to correct the model parameters; if the deviation exceeds the preset threshold, return to Step 2 to re-optimize the BeiDou data preprocessing parameters until the accuracy requirements are met; the ground control points must cover the entire mapped area and be evenly distributed; the accuracy check uses the plane position error and elevation error as evaluation indicators. If the plane position error is >0.5m or the elevation error is >0.3m, the Kalman filter parameters in Step 2 or the feature point matching threshold in Step 4 need to be re-optimized until the accuracy meets the standard.

[0023] Example 2: As Figure 1 , Figure 3 As shown, a 3D real-world map of a forest cover area is constructed by following these steps. Step 1. Airborne Equipment Deployment and BeiDou System Initialization: Install a single BeiDou receiver, laser scanner, IMU, and data storage module on the airborne fixed-wing aircraft platform. Complete time synchronization and parameter calibration between the BeiDou system and the laser equipment, and set the BeiDou positioning sampling frequency and laser scanning resolution. The airborne equipment deployment must meet the BeiDou signal reception requirements. In forested areas, the antenna height and angle of the BeiDou receiver need to be adjusted to ensure that at least four satellites are visible. The scanning resolution of the laser scanner is dynamically set according to the mapping accuracy requirements. A resolution of 0.1-0.5m is used for geological exploration scenarios, and 0.5-1m is used for emergency rescue scenarios to improve data acquisition efficiency. Step 2. BeiDou Positioning Data Preprocessing: Receive raw positioning data from a single BeiDou system (including longitude, latitude, elevation, and timestamp). Use a Kalman filter algorithm to eliminate ionospheric and tropospheric delay errors. Combine this with IMU data to compensate for attitude (roll, pitch, yaw) fluctuations of the airborne platform, obtaining high-precision real-time position and attitude parameters. Error compensation in BeiDou positioning data preprocessing includes: correcting satellite orbit errors using BeiDou ephemeris data; applying an atmospheric delay model to eliminate the influence of the ionosphere and troposphere on the signal; and fusing IMU data with BeiDou positioning data using a 20Hz loosely coupled IMU with a sampling frequency 2-5 times the BeiDou positioning frequency to compensate for instantaneous attitude changes of the airborne platform in real time, obtaining a POS with a horizontal plane ≤0.3 m and an elevation ≤0.2 m. Step 3. Laser Point Cloud Acquisition and Adaptive Denoising: The airborne laser scanner acquires surface point cloud data according to preset parameters. The data includes the three-dimensional coordinates, grayscale values, and reflection intensity of the points. Based on the spatial density and grayscale characteristics of the point cloud, an adaptive threshold denoising algorithm is applied to remove interference points caused by vegetation occlusion, retaining effective topographic and feature point clouds. The threshold setting of the adaptive point cloud denoising algorithm is based on the local density of the point cloud: for forest-covered areas, 0.6 times the average density is used as the threshold. A higher threshold is used to remove interference points caused by vegetation. The denoised data must meet the requirement of an effective point cloud retention rate of ≥90%. If the standard is not met, the denoising threshold is readjusted until the requirement is met. Step 4. Feature Point Extraction and Terrain Matching: The SIFT algorithm is used to extract terrain and feature points from the denoised point cloud. The number of feature points should be no less than 5% of the total number of points. Through a terrain feature constraint matching model, the feature points are associated with the coordinate system of BeiDou positioning to establish a mapping relationship between point cloud data and geographic coordinates. The terrain feature constraint matching model is constructed as follows: using the location of the BeiDou positioning airborne platform as a reference, combined with the scanning angle and distance of the laser scanner, the absolute geographic coordinates (WGS84 coordinate system) of the feature points are calculated to establish the spatial geometric relationship of the point cloud. Through the elevation difference and slope terrain parameters of the feature points, mismatched points caused by scanning angle deviation are eliminated, and the matching accuracy should reach more than 95%. Step 5. Dynamic Coordinate Transformation and 3D Modeling: Based on the WGS84 coordinate system of BeiDou positioning, point cloud data is converted into a target coordinate system (such as the National Geodetic Coordinate System 2000) through a dynamic coordinate transformation mechanism. An irregular triangular mesh modeling method is applied to construct a 3D terrain model, and texture information of ground features is superimposed to generate a 3D real-world map. The dynamic coordinate transformation mechanism is based on preset coordinate transformation parameters (such as four parameters) and dynamically adjusts the transformation coefficients according to the real-time position of BeiDou positioning to ensure seamless stitching of point cloud data from different areas in the target coordinate system. During the 3D modeling process, for dense urban building clusters, building outline texture information needs to be superimposed to improve the visual recognizability of the real-world map. Step 6. Accuracy Check and Model Optimization: Select ground control points with known coordinates (at least 3), calculate the coordinate deviation of the corresponding points in the 3D real-world map, and apply the least squares principle to correct the model parameters; if the deviation exceeds the preset threshold, return to Step 2 to re-optimize the BeiDou data preprocessing parameters until the accuracy requirements are met; the ground control points must cover the entire mapped area and be evenly distributed; the accuracy check uses the plane position error and elevation error as evaluation indicators. If the plane position error is >0.5m or the elevation error is >0.3m, the Kalman filter parameters in Step 2 or the feature point matching threshold in Step 4 need to be re-optimized until the accuracy meets the standard.

[0024] Example 3: As Figure 1 , Figure 4 As shown, construct a 3D real-world map of a mountain canyon using the following steps. Step 1. Airborne Equipment Deployment and BeiDou System Initialization: Install a single BeiDou receiver, laser scanner, IMU, and data storage module on the airborne fixed-wing aircraft platform. Complete time synchronization and parameter calibration between the BeiDou system and the laser equipment, and set the BeiDou positioning sampling frequency and laser scanning resolution. The airborne equipment deployment must meet the BeiDou signal reception requirements. In forested areas or densely populated urban areas, the antenna height and angle of the BeiDou receiver need to be adjusted to ensure that at least four satellites are visible. The scanning resolution of the laser scanner is dynamically set according to the mapping accuracy requirements. A resolution of 0.1-0.5m is used for geological exploration scenarios, and 0.5-1m is used for emergency rescue scenarios to improve data acquisition efficiency. Step 2. BeiDou Positioning Data Preprocessing: Receive raw positioning data from a single BeiDou system (including longitude, latitude, elevation, and timestamp). Use a Kalman filter algorithm to eliminate ionospheric and tropospheric delay errors. Combine this with IMU data to compensate for attitude (roll, pitch, yaw) fluctuations of the airborne platform, obtaining high-precision real-time position and attitude parameters. Error compensation in BeiDou positioning data preprocessing includes: correcting satellite orbit errors using BeiDou ephemeris data; applying an atmospheric delay model to eliminate the influence of the ionosphere and troposphere on the signal; and fusing IMU data with BeiDou positioning data using a 20Hz loosely coupled IMU with a sampling frequency 2-5 times the BeiDou positioning frequency to compensate for instantaneous attitude changes of the airborne platform in real time, obtaining a POS with a horizontal plane ≤0.3 m and an elevation ≤0.2 m. Step 3. Laser Point Cloud Acquisition and Adaptive Denoising: The airborne laser scanner acquires surface point cloud data according to preset parameters. The data includes the three-dimensional coordinates, grayscale values, and reflection intensity of the points. Based on the spatial density and grayscale characteristics of the point cloud, an adaptive threshold denoising algorithm is applied to remove interference points caused by vegetation occlusion and building shadows, retaining effective topographic and feature point clouds. The threshold setting of the adaptive point cloud denoising algorithm is based on the local density of the point cloud: for complex terrain areas such as mountainous canyons, a lower threshold is used to retain detailed points, and the denoised data must meet the requirement of an effective point cloud retention rate ≥90%; for mountainous canyon areas, 0.2 times the average density is used as the threshold to remove isolated interference points; if the standard is not met, the denoising threshold is readjusted until the requirements are met. Step 4. Feature Point Extraction and Terrain Matching: The SIFT algorithm is used to extract terrain feature points (such as mountain peaks, valleys, and cliff edges) and ground feature points from the denoised point cloud. The number of feature points should not be less than 5% of the total number of points. Through a terrain feature constraint matching model, the feature points are associated with the coordinate system of BeiDou positioning to establish a mapping relationship between point cloud data and geographic coordinates. The terrain feature constraint matching model is constructed as follows: using the location of the airborne platform of BeiDou positioning as the reference, combined with the scanning angle and distance of the laser scanner, the absolute geographic coordinates (WGS84 coordinate system) of the feature points are calculated to establish the spatial geometric relationship of the point cloud. Through the elevation difference and slope terrain parameters of the feature points, mismatched points caused by scanning angle deviation are eliminated, and the matching accuracy should reach more than 95%. Step 5. Dynamic Coordinate Transformation and 3D Modeling: Based on the WGS84 coordinate system of BeiDou positioning, point cloud data is converted into a target coordinate system (such as the National Geodetic Coordinate System 2000) through a dynamic coordinate transformation mechanism. An irregular triangular mesh modeling method is applied to construct a 3D terrain model, and texture information of ground features is superimposed to generate a 3D real-world map. The dynamic coordinate transformation mechanism is based on preset coordinate transformation parameters (such as seven parameters) and dynamically adjusts the transformation coefficients according to the real-time position of BeiDou positioning to ensure seamless stitching of point cloud data from different areas in the target coordinate system. During the 3D modeling process, for dense urban building clusters, building outline texture information needs to be superimposed to improve the visual recognizability of the real-world map. Step 6. Accuracy Check and Model Optimization: Select ground control points with known coordinates (at least 3), calculate the coordinate deviation of the corresponding points in the 3D real-world map, and apply the least squares principle to correct the model parameters; if the deviation exceeds the preset threshold, return to Step 2 to re-optimize the BeiDou data preprocessing parameters until the accuracy requirements are met; the ground control points must cover the entire mapped area and be evenly distributed; the accuracy check uses the plane position error and elevation error as evaluation indicators. If the plane position error is >0.5m or the elevation error is >0.3m, the Kalman filter parameters in Step 2 or the feature point matching threshold in Step 4 need to be re-optimized until the accuracy meets the standard.

Claims

1. A method for real-time airborne laser scanning mapping of complex geographical environments using a single BeiDou system-assisted 3D real-scene map, characterized by: Based on the real-time positioning (longitude, latitude, elevation) and timing data of the airborne platform obtained by the single Beidou system, combined with the surface point cloud data collected by the airborne laser scanner, an adaptive point cloud denoising algorithm, terrain feature constraint matching model and dynamic coordinate transformation mechanism are applied to construct a real-time mapping technology system for three-dimensional real scene maps of complex geographical environments. The basic steps include: BeiDou data acquisition and preprocessing - laser point cloud acquisition and denoising - feature point extraction and matching - 3D model construction - accuracy verification and optimization; based on the differences in terrain features of different complex environments (mountains, forests, cities), the algorithm parameters are dynamically adjusted to achieve high-precision real-time output of 3D real-scene maps.

2. The method for real-time airborne laser scanning mapping of complex geographical environments using a single BeiDou system-assisted three-dimensional real-scene map, as described in claim 1, is characterized by: The specific implementation steps of this method are as follows: Step 1. Airborne equipment deployment and BeiDou system initialization: Install a single BeiDou receiver, laser scanner, IMU and data storage module on the airborne platform (UAV, fixed-wing aircraft), complete the time synchronization and parameter calibration of the BeiDou system and laser equipment, and set the BeiDou positioning sampling frequency and laser scanning resolution; Step 2. BeiDou positioning data preprocessing: Receive the raw positioning data from a single BeiDou system, eliminate ionospheric and tropospheric delay errors through Kalman filtering algorithm, and combine IMU data to compensate for the attitude (roll, pitch, yaw) fluctuations of the airborne platform to obtain the high-precision real-time position and attitude parameters of the airborne platform; Step 3. Laser point cloud acquisition and adaptive denoising: The airborne laser scanner acquires surface point cloud data according to preset parameters. Based on the spatial density and grayscale characteristics of the point cloud, an adaptive threshold denoising algorithm is applied to remove interference points caused by vegetation occlusion and building shadows, while retaining effective shape and ground feature point clouds. Step 4. Feature point extraction and terrain matching: Extract terrain feature points (such as mountain tops, valleys, and cliff edges) and ground feature points (such as building corners and road edges) from the denoised point cloud. Through the terrain feature constraint matching model, associate the feature points with the coordinate system of BeiDou positioning to establish the mapping relationship between point cloud data and geographic coordinates. Step 5. Dynamic coordinate transformation and 3D modeling: Based on the WGS84 coordinate system of Beidou positioning, the point cloud data is converted into the target coordinate system (such as the National Geodetic Coordinate System 2000) through a dynamic coordinate transformation mechanism. The irregular triangular mesh modeling method is applied to construct a 3D terrain model, and the texture information of the ground features is superimposed to generate a 3D real-scene map. Step 6. Accuracy Check and Model Optimization: Select ground control points with known coordinates (no fewer than 3), calculate the coordinate deviation of the corresponding points in the 3D real-world map, and apply the least squares principle to correct the model parameters; if the deviation exceeds the preset threshold, return to Step 2 to re-optimize the BeiDou data preprocessing parameters until the accuracy requirements are met.

3. The method for real-time airborne laser scanning mapping of complex geographical environments assisted by a single BeiDou system, as described in claim 2, is characterized by: In Step 1, the deployment of airborne equipment must meet the conditions for receiving BeiDou signals. In forest-covered areas or densely built-up urban areas, the antenna height and angle of the BeiDou receiver need to be adjusted to ensure that at least 4 satellites are visible. The scanning resolution of the laser scanner is dynamically set according to the mapping accuracy requirements. A resolution of 0.1-0.5m is used in geological exploration scenarios, and a resolution of 0.5-1m is used in emergency rescue scenarios to improve acquisition efficiency.

4. The method for real-time airborne laser scanning mapping of complex geographical environments using a single BeiDou system-assisted three-dimensional real-scene map, as described in claim 2, is characterized by: In Step 2, error compensation in BeiDou positioning data preprocessing includes: correcting satellite orbit errors using BeiDou system ephemeris data, and applying atmospheric delay models to eliminate the influence of the ionosphere and troposphere on the signal; the fusion of IMU data and BeiDou positioning data adopts a loose coupling method, with a sampling frequency 2-5 times that of BeiDou positioning frequency, in order to compensate for the instantaneous attitude changes of the airborne platform in real time.

5. The method for real-time airborne laser scanning mapping of complex geographical environments assisted by a single BeiDou system, as described in claim 2, is characterized in that: In Step 3, the threshold setting of the adaptive point cloud denoising algorithm is based on the local density of the point cloud: for complex terrain areas such as mountainous canyons, a lower threshold is used to retain detailed points; for forest-covered areas, a higher threshold is used to remove vegetation interference points; the denoised data must meet the requirement of an effective point cloud retention rate of ≥90%.

6. The method for real-time airborne laser scanning mapping of complex geographical environments assisted by a single BeiDou system, as described in claim 2, is characterized in that: In Step 4, the terrain feature constraint matching model is constructed in the following way: using the location of the airborne platform with Beidou positioning as the reference, the scanning angle and distance of the laser scanner are calculated to establish the spatial geometric relationship of the point cloud; through the elevation difference and slope terrain parameters of the feature points, mismatch points caused by scanning angle deviation are eliminated, and the matching accuracy needs to reach more than 95%.

7. The method for real-time airborne laser scanning mapping of complex geographical environments using a single BeiDou system-assisted three-dimensional real-scene map, as described in claim 2, is characterized by: In Step 5, the dynamic coordinate transformation mechanism dynamically adjusts the transformation coefficients based on preset coordinate transformation parameters (such as seven parameters or four parameters) and the real-time position of Beidou positioning to ensure that point cloud data from different areas are seamlessly stitched together in the target coordinate system. During the 3D modeling process, for dense urban building clusters, it is necessary to overlay building outline texture information to improve the visual recognition of the real-world map.

8. The method for real-time airborne laser scanning mapping of complex geographical environments assisted by a single BeiDou system, as described in claim 2, is characterized in that: In Step 6, the selection of ground control points must cover the entire mapping area and be evenly distributed. The accuracy check uses the mean square error of the plane position and the mean square error of the elevation as evaluation indicators. If the mean square error of the plane is >0.5m or the mean square error of the elevation is >0.3m, the Kalman filter parameters in Step 2 or the feature point matching threshold in Step 4 need to be re-optimized until the accuracy meets the standard.

Images