不同机载LiDAR点云滤波算法对比分析

Comparative Analysis on Different Airborne Li DAR Point Cloud Filtering Algorithms

针对目前各种机载点云滤波算法缺乏详细对比的问题,分别从定性和定量两方面对常用的3种滤波算法进行了对比分析,得出了滤波误差最小的滤波方法,以及各滤波算法对不同地形数据的适应情况。利用国际摄影测量与遥感学会(International Society for Photogrammetry and Remote Sensing,ISPRS)提供的标准滤波数据进行实验。结果表明,6组实验数据中,布料模拟滤波算法的总误差都是最小的;对于平坦地形数据,布料模拟滤波算法最合适,对于地形起伏的复杂地形数据,布料模拟滤波算法和渐进三角网滤波算法总误差值接近,但后者的滤波时间更短。

In view of the lack of detailed comparison of various airborne point cloud filtering algorithms, we compare the three filtering algorithms from qualitative and quantitative aspects to obtain the filtering method with the smallest filtering error and the adaptation of each filtering algorithm to different terrain data. The standard filtering data provided by International Society for Photogrammetry and Remote Sensing(ISPRS)are used in this experiment. The results show that the total error of the cloth simulation filtering algorithm is the smallest among the 6 groups of experimental data. For flat terrain data,the cloth simulation filtering algorithm is the most suitable. For complex terrain data with undulating terrain,the total error values of cloth simulation filtering algorithm and progressive triangulation filtering algorithm are relatively close,but the latter has a shorter filtering time.