A Simplified Method Based on Terrain Complexity for Lidar Point Cloud and Its Uncertainty Analysis

LiDAR is a technology to acquire object surface measurements which integrates GPS, IMU, laser scanning and ranging system and imaging devices together. LIDAR technology has the characteristics of highly automation, short data production cycle, the little effect of external environment and high precision and accuracy to acquire measurement information. But the number of liDAR point cloud is huge. When using large amounts of point cloud data to construct DEM, instead of improving the accuracy of DEM no significant effect, it will lead to the rapid decline in data processing speed. So it is necessary to simplify the LiDAR point cloud. When simplifying the point cloud, the criterions of point cloud simplification directly influence the distribution and quality of retention points. Usually, the point simplification criteria are based on topographic feature. Hence, this paper will proposal a new approach based on terrain complexity metrics to simplify LiDAR point cloud. Terrain complexity index present a comprehensive description of topographic features. First the index is calculated based on the existing rough precision DEM data; next, find out the point cloud simplification threshold according to the index; then set simplify rules to retain the feature points and simplify the useless points; finally, using geostatistical method, high accuracy DEM is constructed by the retention points and the precision and accuracy of LiDAR point cloud simplification is evaluated. The method will be expected to improve the precision and accuracy of LiDAR point cloud simplification. The experiment shows that the proposed simplified method can realize the simplification of LiDAR point cloud, identify terrain feature points effectively, improve the efficiency of the algorithm greatly and help to generate DEMs with a higher precision.