基于小波多分辨率方向高程限差的LiDAR数据滤波

Filter of LIDAR data based on wavelet multi-resolution and directional height tolerance

本文在研究了小波多分辨率分析原理的基础上,将小波变换引入到机载LIDAR数据滤波中,利用金字塔分层理论实现原始DSM数据的快速组织,加快计算速度,提高滤波效率。同时在滤波阈值选择中顾及整个区域的方向高程差,使判断条件更加符合地形特点,提高滤波的准确度。本文最后利用实际的LIDAR数据进行了实验和精度分析。

After studying the principle of wavelet multi-resolution analysis, this paper took wavelet transform into LIDAR data filtering. By making use of pyramid-delaminating theory, the fast organization of raw DSM data came true, while the computing speed increased and the filtering speed improved. At the same time, by considering the directional height tolerance of the whole region in the process of filtering threshold selection, determining factors were made better accordance with topographic characteristics to improve the precision of filtering. At last, experiment was carried out on DSM data produced by LIDAR, and the results proved the validity and ad- vantage of the advanced filtering method.