The Effect Of Meshing Algorithms On DEM Accuracy Of Lidar

Lidar (Light Detection and Ranging) was a new remote sensing technology, whichcould quickly obtain accurate geographic information. Compared with the traditionalmeasurement techniques, Lidar had many advantages, such as initiative, penetrating,non-contact and fast, etc., which made it very important in the field of surveying andmapping applications. Lidar could acquire point cloud data which contained a largenumber of discrete points’ three-dimensional coordinates, the point cloud data didn’tdirectly reflect the terrain or the target’s situation. If you wanted to solve this problem,you must obtain terrain or the target’s DEM through processing the point cloud data withmeshing algorithms. The DEM generated by different meshing algorithms would bedifferent, naturally, DEM accuracy was also different. This article would study the effectof meshing algorithms on DEM accuracy, hoping to find highest precision’s algorithmsin different situation.Firstly, the airborne Lidar system components and a detailed derivation of theairborne LIDAR point cloud data acquisition principle were introduced. Airborne Lidarmainly contained laser range finder, GPS receivers, inertial navigation system. Throughestablishment of the coordinate system, conversion between coordinate systems andbuilding conformational equation，from the measured data,we got the laser foot points’three-dimensional coordinates,namely, the point cloud data. Finally error sourcesimpacting DEM accuracy were analyzed.Then we selected three common meshing algorithms: Rectangular Centre Algorithm,Triangulated Regular Algorithm and Delaunay Triangulation Algorithm, and also describedtheir principle and implementation steps in detail. Then we proposed the evaluation methodof DEM accuracy. Based on ASTER elevation data, with regular and irregular cloudpoints, simulation experiment was carried out to obtain three terrains’ DEM with threemeshing algorithms. Then according to the proposed evaluation method of DEMaccuracy, we calculated and compared the DEM accuracy of the three algorithms, andfound highest precision’s algorithms in different situations.Finally, we used range-gated Lidar to do imaging experiment with a plaster. By therelated processing and a forest, we got the distance images which would replace the pointcloud. We used the distance images to generate DEM，and calculated the DEM accuracy.The final result proved the correctness of the conclusion of the simulation experiment.