A high-precision terrain mapping processing method and system for unmanned aerial vehicle aerial photography

Abstract

The invention discloses a high-precision terrain mapping processing method and system for unmanned aerial vehicle aerial photography, which relates to the technical field of terrain mapping, and comprises a data acquisition module, a three-dimensional terrain reconstruction module, a data processing module and a terrain mapping output module; the data acquisition module is used to obtain terrain data of a mapping area, the terrain data comprising: point cloud data and position data of the unmanned aerial vehicle during flight, and simultaneously obtains terrain data processed by the data processing module, performs negative feedback regulation on the acquisition of terrain data, and transmits the processed terrain data to the three-dimensional terrain reconstruction module; the three-dimensional terrain reconstruction module is used to remove noise points and invalid points in the point cloud data, and simultaneously processes the same feature points between different point cloud data, calculates the initial position and posture of a LiDAR radar device of each frame of point cloud data, generates a dense point cloud, and transmits the dense point cloud to the data processing module.

Description

Technical Field

[0001] The present invention relates to the technical field of terrain surveying and mapping, and in particular to a high-precision terrain surveying and mapping processing method and system for unmanned aerial vehicle aerial photography. Background Art

[0002] In the mid-19th century, with the emergence of photography technology, photogrammetry began to be used in topographic mapping. Through photos taken from the air or on the ground, stereo models were created and topographic information was extracted. At the end of the 19th century, countries began to formulate standards and specifications for topographic maps, making the production of topographic maps more systematic and standardized. With the invention of the airplane, aerial photogrammetry developed rapidly, allowing large-scale topographic data to be obtained from the air, greatly improving mapping efficiency. After entering the 21st century, drones have become an important tool for topographic mapping. Drones carry various sensors and can collect topographic data at a higher resolution in a smaller range. The development of topographic mapping processing methods is the result of multidisciplinary cross-disciplinary research, involving mathematics, physics, engineering, computer science and other fields.

[0003] However, the accuracy of traditional topographic mapping on slopes changes with the terrain, making the density of collected data uneven, thus affecting the accuracy. At a certain flight altitude, for terrain with high and low drops, there is a certain deviation in the collected data, which may cause a certain degree of distortion in the generated topographic mapping model.

[0004] In view of the above technical defects, a solution is now proposed. Summary of the invention

[0005] The purpose of the present invention is to provide a high-precision terrain mapping and processing system for drone aerial photography to solve the above-mentioned technical defects. The present invention plans the terrain mapping route by arranging several ground control points with known coordinates in the mapping area through obtaining the mapping area, objectives and requirements, obtaining regional terrain point cloud data, drone altitude and terrain altitude of unmapped areas, and judging and adjusting the vertical flight altitude of the drone and the mapping area in real time. After processing the point cloud data and dense point cloud, the model is generated.

[0006] The purpose of the present invention can be achieved by the following technical solutions: A high-precision terrain mapping processing system for drone aerial photography, comprising a data acquisition module, a three-dimensional terrain reconstruction module, a data processing module and a terrain mapping output module;

[0007] The data acquisition module is used to obtain the terrain data of the surveying area. The terrain data includes: point cloud data and the position data of the UAV during flight. At the same time, it obtains the terrain data processed by the data processing module, performs negative feedback regulation on the acquisition of terrain data, and transmits the processed terrain data to the 3D terrain reconstruction module;

[0008] The 3D terrain reconstruction module is used to remove noise points and invalid points in the point cloud data, process the same feature points between different point cloud data, calculate the initial position and posture of the LiDAR radar device for each frame of point cloud data, generate a dense point cloud, and transmit the dense point cloud to the data processing module;

[0009] The data processing module is used to process the point cloud data obtained by the drone, obtain the drone's flight altitude in real time according to the point cloud density requirements, plan the flight altitude of the unmapped area on the drone's flight path, and send the planning results to the data acquisition module and the terrain mapping output module;

[0010] The terrain mapping output module is used to obtain dense point clouds, map the original images to three-dimensional models, and output the generated three-dimensional models.

[0011] Preferably, the process of acquiring the terrain data of the surveying and mapping area and the process of acquiring the original data of the surveying and mapping area are as follows:

[0012] By obtaining the survey area, objectives and requirements, several ground control points with known coordinates are arranged in the survey area, and the terrain survey route is planned through software;

[0013] Conduct survey and detection of the surveying area, and use the line segmentation and flight operation mode to obtain the original data of the surveying area, including: terrain characteristics, vegetation coverage, geological structure data, and the height difference of the terrain ups and downs in the surveying area;

[0014] Line segmentation: The entire survey area is divided into several sub-areas. Each sub-area is executed as an independent flight mission, and the original data of the survey area is downloaded regularly;

[0015] Operation in batches: In surveying and mapping tasks, the flight mission is divided into multiple batches to complete.

[0016] Preferably, the steps of arranging and maintaining ground control points in the target area are:

[0017] S101: Prepare high-precision GNSS receivers, survey poles, marking materials and recording tools; ground control points are evenly distributed in the survey area, at least 4 ground control points are required, fixed vertically and covering different parts of the entire survey area;

[0018] S102: Record the coordinates, number and surrounding environment characteristics of each ground control point;

[0019] S103: Exporting the collected ground control point coordinate data;

[0020] S104: Check the accuracy of the control point coordinates and regularly visit the control points to see if they are damaged or moved;

[0021] Arrange ground control points in the surveying area, obtain the original data of the surveying area, and set the flight parameters according to the actual situation of the surveying area and the operation requirements.

[0022] Preferably, the process of acquiring regional terrain point cloud data and drone altitude is as follows:

[0023] The LiDAR radar device suspended at the bottom of the drone emits laser pulses and measures the time required for the laser pulses to be reflected from the terrain surface back to the sensor, calculates the distance of each laser pulse from the terrain surface, and collects point cloud data;

[0024] By collecting a large amount of point cloud data, the LiDAR radar device emits a laser pulse perpendicular to the ground and measures the time required for the laser pulse to return to the sensor to calculate the distance, determine the flight altitude of the drone, and adjust the flight altitude according to the required ground point cloud density and point cloud density. Point cloud density refers to the number of points per unit area. The formula is:

[0025] Point cloud density = (flight speed × flight altitude) / (scanning frequency × number of scan lines)

[0026] Ground point cloud density: the actual distance represented by each laser point on the ground;

[0027] Scanning frequency: the number of laser pulses emitted by the LiDAR radar device per second;

[0028] Scanning lines: the number of lines scanned per second by the LiDAR radar device;

[0029] Flight speed: the speed of the drone when performing a mission;

[0030] Flight altitude: The vertical height of the drone relative to the survey area.

[0031] Preferably, in the surveying area, the vertical flight height range of the drone is obtained:

[0032] Adjustments are made based on the terrain undulations in the surveyed area. The density of the terrain point cloud in the surveyed area is uniform. The minimum flight altitude of the drone is calculated based on the original data of the surveyed area, that is, the minimum safe flight altitude of the drone.

[0033] By obtaining the required point cloud density requirements of the surveying area and the maximum scanning frequency of the LiDAR radar equipment, the maximum flight altitude of the UAV is calculated, that is, the maximum altitude required for the UAV point cloud density collection;

[0034] When the UAV is at a minimum flight altitude and below the maximum flight altitude during flight, the altitude is the UAV flight data collection interval. At the same time, the position coordinates of the ground control point are obtained, and the position parameters at this time are corrected and adjusted. Based on the position parameters, the height difference of the terrain ups and downs in the surveying area is obtained;

[0035] When the drone's flight altitude is lower than the minimum flight altitude but higher than the maximum flight altitude, the drone's flight altitude needs to be adjusted urgently.

[0036] Preferably, the process of obtaining the terrain height of the unmapped area on the flight path of the drone is as follows:

[0037] The UAV maintains a horizontal flight attitude. The LiDAR radar device emits low-frequency laser pulses to the unmapped areas on the UAV's path and measures the time required for the laser pulses to be reflected from the terrain surface back to the sensor. It calculates the distance between the UAV and the terrain surface of the unmapped area, scans the height difference of the unmapped terrain, and records the angle between the LiDAR radar device and the horizontal plane of the UAV when the laser pulse is emitted. By obtaining the height of the UAV and the angle between the LiDAR radar device and the horizontal plane when the laser pulse is emitted, as well as the distance between the UAV and the terrain surface of the unmapped area, the cosine theorem is used to calculate the vertical height between the unmapped area and the UAV, and the UAV's flight altitude is adjusted in time to obtain point cloud data with uniform density.

[0038] Preferably, the process of processing point cloud data and generating dense point cloud is as follows:

[0039] Remove noise points and invalid points in the point cloud data, extract key points, find the same feature points between different point cloud data, calculate the initial position and posture of the LiDAR radar device in each frame of point cloud data through feature matching results, overlap adjacent point cloud data, obtain the feature points of each point cloud data and the corresponding points in the surrounding point cloud data, use the stereo matching algorithm to generate a depth map, convert the depth map of each point cloud data into a common reference coordinate system, check the consistency between the depth maps generated by different point cloud data, and use weighted average to merge multiple depth maps into a dense point cloud.

[0040] Preferably, the vertical flight height of the UAV and the surveying area is determined and adjusted as follows:

[0041] By acquiring the real-time location data of the UAV and receiving the signal from the ground control point, the flight path of the UAV is corrected so that the UAV can fly and map according to the planned path;

[0042] The height of the terrain in the unmapped area on the flight path of the drone is calculated in real time. By obtaining the distance between the drone and the terrain surface of the unmapped area and the vertical flight height between the drone and the mapped area at this time, the flight height adjustment judgment is made:

[0043] When the terrain height of the unsurveyed area is greater than or equal to the vertical flight height of the drone and the surveyed area at this time, the drone's flight altitude is lowered, while keeping the drone's flight altitude greater than or equal to the drone's safe flight altitude;

[0044] When the terrain height of the unsurveyed area is lower than the vertical flight height between the drone and the surveyed area, increase the drone's flight height while keeping the drone's flight height lower than or equal to the maximum height required for drone point cloud density collection;

[0045] Safe flight altitude of drones: By surveying and detecting the surveying area, the maximum height of the objects in the surveying area is obtained, which is the safe flight altitude of the drones;

[0046] Maximum altitude required for drone point cloud density collection: Get the required point cloud density and calculate the maximum altitude required for drone point cloud density collection.

[0047] Preferably, the process of mapping the texture map to the triangle mesh is as follows:

[0048] By covering the original images of the surveying area, selecting the original images under clear lighting conditions, performing distortion correction on the original images, generating texture coordinates for each triangle vertex, fusing the texture information in multiple original images together, generating the final texture map, mapping the generated texture map with the texture coordinates of the triangular mesh, and mapping the texture map to each triangle of the triangular mesh.

[0049] A high-precision terrain mapping processing method for unmanned aerial vehicle aerial photography, comprising:

[0050] Step 1: Obtain the terrain data of the survey area, arrange several ground control points with known coordinates in the survey area, obtain the survey area, target and requirements, obtain point cloud data through the LiDAR radar equipment suspended at the bottom of the drone and the planned terrain survey route, and transmit the terrain data to step 2;

[0051] Step 2: Acquire point cloud data and remove noise points and invalid points in the point cloud data; match the same feature points between different point cloud data, calculate the initial position and posture of the LiDAR radar device for each frame of point cloud data, overlap adjacent point cloud data, fuse them into dense point cloud, and pass the dense point cloud data to step 3;

[0052] Step 3: Process the dense point cloud data obtained by the drone, obtain the drone's flight altitude in real time according to the point cloud density requirements, make a flight altitude plan for the unsurveyed terrain on the drone's flight path, adjust the drone's flight in real time, and pass the processed dense point cloud data and flight altitude plan to step 4;

[0053] Step 4: Obtain the original image, map the original image to the three-dimensional model, fuse the texture information in multiple original images together, generate a texture map, and output the generated three-dimensional model.

[0054] The beneficial effects of the present invention are as follows:

[0055] The present invention arranges a number of ground control points with known coordinates in the surveying and mapping area to correct geometric errors and geographical location information in topographic mapping, improves the spatial positioning accuracy in the process of topographic mapping, obtains the height difference of the terrain ups and downs in the surveying and mapping area, adopts a line segmentation and flight operation mode, and plans topographic mapping routes to ensure that there is enough overlap between point cloud data to support subsequent terrain data processing.

[0056] The present invention acquires the terrain height of the unsurveyed area on the UAV flight path and the real-time height of the current position, performs negative feedback regulation on the collection of terrain data, and adjusts the flight altitude in real time according to the terrain on the path to ensure that the terrain point cloud density of the entire surveying area is uniform and meets the safe flight conditions. BRIEF DESCRIPTION OF THE DRAWINGS

[0057] The present invention will be further described below in conjunction with the accompanying drawings;

[0058] Figure 1 It is a flowchart of the system of the present invention;

[0059] Figure 2 It is a reference diagram of the method of the present invention. DETAILED DESCRIPTION

[0060] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention. Example 1

[0061] See also Figure 1 As shown, this embodiment provides a high-precision terrain mapping processing system for drone aerial photography, including the following modules:

[0062] The data acquisition module is used to obtain the terrain data of the surveying area. The terrain data includes: point cloud data and the position data of the UAV during flight. At the same time, it obtains the terrain data processed by the data processing module, performs negative feedback regulation on the acquisition of terrain data, and transmits the processed terrain data to the 3D terrain reconstruction module;

[0063] By obtaining the survey area, objectives and requirements, several ground control points with known coordinates are arranged in the survey area to correct the geometric errors and geographical location information of topographic surveying, improve the spatial positioning accuracy during topographic surveying, and plan the topographic surveying route through software to ensure that there is enough overlap between point cloud data to support subsequent topographic data processing;

[0064] Acquisition of original data of surveying and mapping area:

[0065] Conduct surveys and detections in the surveying and mapping area to obtain the original data of the surveying and mapping area, including data on terrain features, vegetation coverage, and geological structure, and check the height difference of the terrain in the surveying and mapping area, using line segmentation and flight operation mode to obtain;

[0066] Line segmentation: Divide the entire survey area into several sub-areas or "segments", each of which is executed as an independent flight mission, which can more effectively manage and monitor the flight missions of each small area. The amount of data in each small segment is relatively small, which is convenient for post-processing and quality control;

[0067] Flight-by-flight operation: In surveying and mapping tasks, the flight mission is divided into multiple flights to complete. The battery life of the drone is limited and it is impossible to complete the flight of the entire area at one time. The storage device on the drone has limited capacity and data needs to be downloaded regularly;

[0068] Topographic features: understand the slope, aspect and steepness of the survey area;

[0069] Geological structure: Different types of soil may affect the reflection characteristics of LiDAR signals;

[0070] Vegetation coverage: Dense vegetation may affect the quality of LiDAR data, and different tree species have different effects on laser penetration;

[0071] Steps for arranging and maintaining ground control points in the target area:

[0072] S101: Prepare high-precision GNSS receivers, survey poles, marking materials and recording tools; Ground control points are evenly distributed in the survey area, covering different parts of the entire survey area. At least 4 ground control points are required, but for large areas or complex terrain, more ground control points are required to ensure overall accuracy; Choose a location that is easy to identify and not easy to move, such as road intersections, building corners, etc., and stay away from trees, buildings or other objects that may block the signal;

[0073] S102: Record the coordinates, number and surrounding environment characteristics of each ground control point in detail;

[0074] S103: exporting the collected ground control point coordinate data and performing a preliminary check to ensure that there are no obvious errors;

[0075] S104: Perform quality checks on the coordinates of the control points to ensure that they meet the accuracy requirements, and revisit the control points regularly to check whether they are damaged or moved;

[0076] Arrange ground control points in the surveying area, obtain the original data of the surveying area, set the flight parameters, determine the point cloud density requirements of the surveying area, and set them according to the actual situation of the surveying area and the operation requirements;

[0077] Obtain spatial positioning accuracy requirements, set the UAV flight altitude, and obtain regional terrain point cloud data and UAV altitude through LiDAR radar equipment;

[0078] The LiDAR radar device suspended at the bottom of the drone emits laser pulses, measures the time required for the laser pulses to reflect from the terrain surface back to the sensor, and calculates the distance of each laser pulse from the terrain surface to collect point cloud data. By collecting a large amount of point cloud data, a detailed three-dimensional model is created. At the same time, the LiDAR radar device emits laser pulses perpendicular to the ground, and measures the time required for the laser pulses to return to the sensor to calculate the distance, and then determine the flight altitude of the drone. The appropriate flight altitude is determined based on the required ground point cloud density and point cloud density. The ground point cloud density refers to the actual distance represented by each laser point on the ground, and the point cloud density refers to the number of points per unit area. The formula is:

[0079] Point cloud density = (flight speed × flight altitude) / (scanning frequency × number of scan lines)

[0080] Scanning frequency: the number of laser pulses emitted by the LiDAR radar device per second, usually in "points / second". The higher the scanning frequency, the more points are collected per second, thereby increasing the density of the point cloud; Number of scanning lines: the number of lines scanned by the LiDAR radar device per second. The more scanning lines there are, the wider the area covered per second, and the density of the point cloud data will increase accordingly; Flight speed: the speed of the drone when performing a mission; Flight altitude: the vertical height of the drone relative to the surveying area;

[0081] Surveying area, UAV vertical flight altitude range acquisition:

[0082] Adjustments are made according to the terrain undulations of the surveying area to ensure that the terrain point cloud density of the entire surveying area is consistent. The minimum flight altitude of the drone, i.e., the minimum altitude at which the drone can safely fly, is calculated based on the original data of the surveying area. The maximum flight altitude of the drone, i.e., the maximum altitude required for drone point cloud density collection, is calculated by obtaining the required point cloud density requirements of the surveying area and the maximum scanning frequency of the LiDAR radar equipment.

[0083] When the drone's flight altitude is lower than the minimum flight altitude but higher than the maximum flight altitude, the drone's flight altitude needs to be adjusted urgently; when the drone's flight altitude is higher than or equal to the minimum flight altitude but lower than or equal to the maximum flight altitude, it is the height of the drone's flight data collection interval. At the same time, the position coordinates of the ground control points are obtained, and the position parameters are corrected and adjusted in time. According to the position parameters, the height difference of the terrain ups and downs in the surveying and mapping area is obtained to generate a smoother, more natural surface model without obvious gaps, reduce data deviations caused by sparse local point clouds, and ensure the consistency of details in the entire scanning area;

[0084] Acquisition of terrain height in unmapped areas along the flight path of the drone:

[0085] Keep the drone flying in a horizontal posture. The LiDAR radar device emits low-frequency laser pulses to the unmapped area on the drone's path and measures the time required for the laser pulse to be reflected from the terrain surface back to the sensor, thereby calculating the distance between the drone and the terrain surface of the unmapped area. The height difference of the unmapped terrain is continuously scanned, and the angle between the LiDAR radar device and the drone's horizontal plane is recorded. By obtaining the drone's altitude and the angle between the LiDAR radar device and the drone's horizontal plane when the laser pulse is emitted, as well as the distance between the drone and the terrain surface of the unmapped area, the vertical height between the unmapped area and the drone is calculated using the cosine theorem, thereby adjusting the drone's flight altitude in a timely manner to make the density of the point cloud data in the obtained mapping area uniform.

[0086] The 3D terrain reconstruction module is used to remove noise points and invalid points in the point cloud data. At the same time, for the same feature points between different point cloud data, it calculates the initial position and posture of the LiDAR radar device in each frame of point cloud data, generates a dense point cloud, and transmits the dense point cloud to the data processing module;

[0087] Obtain the point cloud data and open source libraries collected by the LiDAR radar device, provide various point cloud processing algorithms, remove noise points and invalid points in the point cloud data, extract key points from the point cloud data, and find the same feature points between different point cloud data. Through the feature matching results, calculate the initial position and posture of the LiDAR radar device for each frame of point cloud data, use the Bundle Adjustment method to optimize the posture parameters of the LiDAR radar device to improve the accuracy, overlap adjacent point cloud data, fuse them into dense point clouds, convert dense point clouds into triangular mesh models, extract texture information from the original image, and map the texture information to the triangular mesh model to make the model more realistic;

[0088] Remove noise points and invalid points from point cloud data. The specific steps are as follows:

[0089] Inaccurate or abnormal points caused by LiDAR radar equipment errors, environmental interference, etc. usually refer to those points that do not belong to the target object or terrain surface, such as points in the sky, points under the ground, or other irrelevant background points. The distance distribution between each point and its neighboring points is calculated to determine whether it is a noise point. The average distance from each point to its neighboring points is calculated by the mean distance method. If the distance deviates significantly from the overall average, it is considered a noise point.

[0090] The code formula is:

[0091] #include<pcl / point_cloud.h>

[0092] #include<pcl / io / pcd_io.h>

[0093] #include<pcl / filters / statistical_outlier_removal.h>

[0094] int main()

[0095] {

[0096] / / Load point cloud data

[0097] pcl::PointCloud <pcl::pointxyz>::Ptrcloud(new pcl::PointCloud <pcl::pointxyz>);

[0098] pcl::io::loadPCDFile <pcl::pointxyz>("input.pcd", *cloud);

[0099] / / Create a statistical filter object

[0100] pcl::StatisticalOutlierRemoval <pcl::pointxyz>sor;

[0101] sor.setInputCloud(cloud);

[0102] sor.setMeanK(50); / / Select the number of neighboring points

[0103] sor.setStddevMulThresh(1.0); / / Standard deviation multiple threshold

[0104] / / Apply filtering

[0105] pcl::PointCloud <pcl::pointxyz>::Ptr cloud_filtered(new pcl::PointCloud <pcl::pointxyz>);

[0106] sor.filter(*cloud_filtered);

[0107] / / Save the filtered point cloud

[0108] pcl::io::savePCDFileASCII("filtered.pcd", *cloud_filtered);

[0109] return 0;

[0110] }

[0111] The method of finding the same feature points between different point cloud data and matching the results, and calculating the initial position and posture of the LiDAR radar device for each frame of point cloud data is as follows:

[0112] Use SIFT 3D to find scale-invariant key points on three-dimensional point cloud data and generate a descriptor for each key point. The descriptor is a compact representation of the local structure around the key point, which is used to compare the similarity between different point cloud data. Use the descriptor to match the key points in different point cloud data, and use the distance metric to evaluate the similarity between two descriptors. After matching, apply the RANSAC algorithm to eliminate false matches, align the two sets of point cloud data through the transformation model, optimize the matching results, and ensure the accuracy and consistency of the matching. Use the initial transformation obtained by rough registration as the starting point, use the ICP algorithm to further optimize the transformation parameters, check the quality of the registration results, and continuously update the position and posture of the LiDAR device in the SLAM system;

[0113] The process of overlapping adjacent point cloud data and fusing them into a dense point cloud is as follows:

[0114] Get the feature points of each point cloud data and the corresponding points in the surrounding point cloud data, use the stereo matching algorithm to estimate the depth value of each point cloud data, generate a depth map, convert the depth map of each point cloud data to a common reference coordinate system, check the consistency between the depth maps generated by different point cloud data, and use weighted average to merge multiple depth maps into a dense point cloud;

[0115] The data processing module is used to process the point cloud data obtained by the drone, obtain the drone's flight altitude in real time according to the point cloud density requirements, plan the flight altitude of the unmapped area on the drone's flight path, and send the planning results to the data acquisition module and the terrain mapping output module;

[0116] By obtaining the real-time location data of the UAV, the flight path of the UAV is guided, and the flight path of the UAV is corrected through the ground control points, so that the UAV can fly and map according to the planned path;

[0117] The height of the terrain in the unmapped area on the flight path of the drone is calculated in real time. By obtaining the distance between the drone and the terrain surface of the unmapped area and the vertical flight height between the drone and the mapped area at this time, the flight height adjustment judgment is made:

[0118] When the terrain height of the unsurveyed area is greater than or equal to the vertical flight height of the drone and the surveyed area at this time, the drone's flight altitude is lowered, while keeping the drone's flight altitude greater than or equal to the drone's safe flight altitude;

[0119] When the terrain height of the unsurveyed area is lower than the vertical flight height between the drone and the surveyed area, increase the drone's flight height while keeping the drone's flight height lower than or equal to the maximum height required for drone point cloud density collection;

[0120] Safe flight altitude of drones: By surveying and detecting the surveying area, the maximum height of the objects in the surveying area is obtained, which is the safe flight altitude of the drones;

[0121] Maximum altitude required for drone point cloud density collection: By substituting the required point cloud density, the maximum altitude required for drone point cloud density collection is calculated;

[0122] The topographic mapping output module is used to obtain dense point clouds, map the original image to a 3D model, and output the generated 3D model;

[0123] By selecting the original image that can cover the best viewing angle of the surveying area, selecting the original image under clear lighting conditions, performing distortion correction on the original image, eliminating the influence of lens distortion, and generating texture coordinates for each triangle vertex so as to map the texture in the original image to the triangular mesh, the texture information in multiple original images is fused together to generate the final texture map, mapping the generated texture map with the texture coordinates of the triangular mesh, and mapping the texture map to each triangle of the triangular mesh, the model has a realistic appearance. Example 2

[0124] See also Figure 2 As shown, based on Example 1, this embodiment also provides a high-precision terrain mapping processing method for drone aerial photography, including the following specific steps:

[0125] Step 1: Obtain the terrain data of the survey area, arrange several ground control points with known coordinates in the survey area, obtain the survey area, target and requirements, obtain point cloud data through the LiDAR radar equipment suspended at the bottom of the drone and the planned terrain survey route, and transmit the terrain data to step 2;

[0126] Step 2: Acquire point cloud data and remove noise points and invalid points in the point cloud data; match the same feature points between different point cloud data, calculate the initial position and posture of the LiDAR radar device for each frame of point cloud data, overlap adjacent point cloud data, fuse them into dense point cloud, and pass the dense point cloud data to step 3;

[0127] Step 3: Process the dense point cloud data obtained by the drone, obtain the drone's flight altitude in real time according to the point cloud density requirements, make a flight altitude plan for the unsurveyed terrain on the drone's flight path, adjust the drone's flight in real time, and pass the processed dense point cloud data and flight altitude plan to step 4;

[0128] Step 4: Obtain the original image, map the original image to the three-dimensional model, fuse the texture information in multiple original images together, generate a texture map, and output the generated three-dimensional model.

[0129] The above embodiments may be implemented in whole or in part by software, hardware, firmware or any other combination thereof. When implemented using software, the above embodiments may be implemented in whole or in part in the form of a computer program product. A person of ordinary skill in the art may appreciate that the units and algorithm steps of each example described in conjunction with the embodiments disclosed herein may be implemented in electronic hardware, or in a combination of computer software and electronic hardware. Whether these functions are performed in hardware or software depends on the specific application and design constraints of the technical solution.

[0130] The units described as separate components may or may not be physically separated, and the components shown as units may or may not be physical units, that is, they may be located in one place or distributed on multiple network units. Some or all of the units may be selected according to actual needs to achieve the purpose of the solution of this embodiment.

[0131] The above description is only a specific implementation manner of the present application, but the protection scope of the present application is not limited thereto. Any technician familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed in the present application, which should be included in the protection scope of the present application.< / pcl::pointxyz> < / pcl::pointxyz> < / pcl::pointxyz> < / pcl::pointxyz> < / pcl::pointxyz> < / pcl::pointxyz>

Claims

1. A high-precision terrain mapping processing system for drone aerial photography, characterized in that: It includes data acquisition module, 3D terrain reconstruction module, data processing module and terrain mapping output module; The data acquisition module is used to obtain the terrain data of the surveying area. The terrain data includes: point cloud data and the position data of the UAV during flight. At the same time, it obtains the terrain data processed by the data processing module, performs negative feedback regulation on the acquisition of terrain data, and transmits the processed terrain data to the 3D terrain reconstruction module; The 3D terrain reconstruction module is used to remove noise points and invalid points in the point cloud data, process the same feature points between different point cloud data, calculate the initial position and posture of the LiDAR radar device for each frame of point cloud data, generate a dense point cloud, and transmit the dense point cloud to the data processing module; The data processing module is used to process the point cloud data obtained by the drone, obtain the drone's flight altitude in real time according to the point cloud density requirements, plan the flight altitude of the unmapped area on the drone's flight path, and send the planning results to the data acquisition module and the terrain mapping output module; The topographic mapping output module is used to obtain dense point clouds, map the original image to a 3D model, and output the generated 3D model; Among them, obtaining the topographic data of the surveying area includes: By obtaining the survey area, objectives and requirements, several ground control points with known coordinates are arranged in the survey area, and the terrain survey route is planned through software; and several ground control points with known coordinates are arranged in the survey area, including: S101: Prepare high-precision GNSS receivers, survey poles, marking materials and recording tools; ground control points are evenly distributed in the survey area, at least 4 ground control points are required, fixed vertically and covering different parts of the entire survey area; S102: Record the coordinates, number and surrounding environment characteristics of each ground control point; S103: Exporting the collected ground control point coordinate data; S104: Check the accuracy of the control point coordinates and regularly visit the control points to see if they are damaged or moved; Arrange ground control points in the surveying area, obtain the original data of the surveying area, and set the flight parameters according to the actual situation of the surveying area and the operation requirements; Conduct survey and detection of the surveying area, and use the line segmentation and flight operation mode to obtain the original data of the surveying area, including: terrain characteristics, vegetation coverage, geological structure data, and the height difference of the terrain ups and downs in the surveying area; Line segmentation: The entire survey area is divided into several sub-areas. Each sub-area is executed as an independent flight mission, and the original data of the survey area is downloaded regularly; Operation in batches: In surveying and mapping tasks, the flight mission is divided into multiple batches to complete.

2. The high-precision terrain mapping processing system for drone aerial photography according to claim 1 is characterized in that: The process of acquiring regional terrain point cloud data and drone altitude is as follows: The LiDAR radar device suspended at the bottom of the drone emits laser pulses and measures the time required for the laser pulses to be reflected from the terrain surface back to the sensor, calculates the distance of each laser pulse from the terrain surface, and collects point cloud data; By collecting a large amount of point cloud data, the LiDAR radar device emits a laser pulse perpendicular to the ground and measures the time required for the laser pulse to return to the sensor to calculate the distance, determine the flight altitude of the drone, and adjust the flight altitude according to the required ground point cloud density and point cloud density. Point cloud density refers to the number of points per unit area. The formula is: Point cloud density = (flight speed × flight altitude) / (scanning frequency × number of scan lines) Ground point cloud density: the actual distance represented by each laser point on the ground; Scanning frequency: the number of laser pulses emitted by the LiDAR radar device per second; Scanning lines: the number of lines scanned per second by the LiDAR radar device; Flight speed: the speed of the drone when performing a mission; Flight altitude: The vertical height of the drone relative to the survey area.

3. The high-precision terrain mapping processing system for drone aerial photography according to claim 2 is characterized in that: Surveying area, UAV vertical flight altitude range acquisition: Adjustments are made based on the terrain undulations in the surveyed area. The density of the terrain point cloud in the surveyed area is uniform. The minimum flight altitude of the drone is calculated based on the original data of the surveyed area, that is, the minimum safe flight altitude of the drone. By obtaining the required point cloud density requirements of the surveying area and the maximum scanning frequency of the LiDAR radar equipment, the maximum flight altitude of the UAV is calculated, that is, the maximum altitude required for the UAV point cloud density collection; When the UAV is at a minimum flight altitude and less than or equal to the maximum flight altitude during flight, it is the height of the UAV flight data collection interval. At the same time, the position coordinates of the ground control point are obtained, the position parameters are corrected and adjusted, and the height difference of the terrain ups and downs in the surveying area is obtained according to the position parameters; When the drone's flight altitude is lower than the minimum flight altitude but higher than the maximum flight altitude, the drone's flight altitude needs to be adjusted urgently.

4. The high-precision terrain mapping processing system for drone aerial photography according to claim 3 is characterized in that: The process of obtaining the terrain height of the unmapped area on the UAV flight path is as follows: The UAV maintains a horizontal flight attitude. The LiDAR radar device emits low-frequency laser pulses to the unmapped areas on the UAV's path and measures the time required for the laser pulses to be reflected from the terrain surface back to the sensor. It calculates the distance between the UAV and the terrain surface of the unmapped area, scans the height difference of the unmapped terrain, and records the angle between the LiDAR radar device and the horizontal plane of the UAV when the laser pulse is emitted. By obtaining the height of the UAV and the angle between the LiDAR radar device and the horizontal plane when the laser pulse is emitted, as well as the distance between the UAV and the terrain surface of the unmapped area, the cosine theorem is used to calculate the vertical height between the unmapped area and the UAV, and the UAV's flight altitude is adjusted in time to obtain point cloud data with uniform density.

5. The high-precision terrain mapping processing system for drone aerial photography according to claim 1 is characterized in that: The process of point cloud data processing and dense point cloud generation is as follows: Remove noise points and invalid points in the point cloud data, extract key points, find the same feature points between different point cloud data, calculate the initial position and posture of the LiDAR radar device in each frame of point cloud data through feature matching results, overlap adjacent point cloud data, obtain the feature points of each point cloud data and the corresponding points in the surrounding point cloud data, use the stereo matching algorithm to generate a depth map, convert the depth map of each point cloud data into a common reference coordinate system, check the consistency between the depth maps generated by different point cloud data, and use weighted average to merge multiple depth maps into a dense point cloud.

6. The high-precision terrain mapping processing system for drone aerial photography according to claim 5 is characterized in that: The vertical flight height between the drone and the surveying area is determined and adjusted as follows: By acquiring the real-time location data of the UAV and receiving the signal from the ground control point, the flight path of the UAV is corrected so that the UAV can fly and map according to the planned path; The height of the terrain in the unmapped area on the flight path of the drone is calculated in real time. By obtaining the distance between the drone and the terrain surface of the unmapped area, as well as the vertical flight height between the drone and the mapped area, the flight height adjustment judgment is made: When the terrain height of the unsurveyed area is greater than or equal to the vertical flight height of the drone and the surveyed area at this time, the drone's flight altitude is lowered, while keeping the drone's flight altitude greater than or equal to the drone's safe flight altitude; When the terrain height of the unsurveyed area is lower than the vertical flight height between the drone and the surveyed area, increase the drone's flight height while keeping the drone's flight height lower than or equal to the maximum height required for drone point cloud density collection; Safe flight altitude of drones: By surveying and detecting the surveying area, the maximum height of the objects in the surveying area is obtained, which is the safe flight altitude of the drones; Maximum altitude required for drone point cloud density collection: Get the required point cloud density and calculate the maximum altitude required for drone point cloud density collection.

7. The high-precision terrain mapping processing system for drone aerial photography according to claim 1 is characterized in that: The process of mapping a texture map to a triangle mesh is as follows: By covering the original images of the surveying area, selecting the original images under clear lighting conditions, performing distortion correction on the original images, generating texture coordinates for each triangle vertex, fusing the texture information in multiple original images together, generating the final texture map, mapping the generated texture map with the texture coordinates of the triangular mesh, and mapping the texture map to each triangle of the triangular mesh.

8. A method for high-precision terrain mapping processing of drone aerial photography, the method being applied to a high-precision terrain mapping processing system for drone aerial photography as claimed in any one of claims 1 to 7, characterized in that: The following steps are involved: Step 1: Obtain the terrain data of the survey area, arrange several ground control points with known coordinates in the survey area, obtain the survey area, target and requirements, obtain point cloud data through the LiDAR radar equipment suspended at the bottom of the drone and the planned terrain survey route, and transmit the terrain data to step 2; Step 2: Acquire point cloud data and remove noise points and invalid points in the point cloud data; match the same feature points between different point cloud data, calculate the initial position and posture of the LiDAR radar device for each frame of point cloud data, overlap adjacent point cloud data, fuse them into dense point cloud, and pass the dense point cloud data to step 3; Step 3: Process the dense point cloud data obtained by the drone, obtain the drone's flight altitude in real time according to the point cloud density requirements, make a flight altitude plan for the unsurveyed terrain on the drone's flight path, adjust the drone's flight in real time, and pass the processed dense point cloud data and flight altitude plan to step 4; Step 4: Obtain the original image, map the original image to the three-dimensional model, fuse the texture information in multiple original images together, generate a texture map, and output the generated three-dimensional model.

Images