针对现有机载LiDAR(light detection and ranging)点云滤波方法在地形起伏剧烈的林区适用性不足的问题,提出一种多分辨率层次布料模拟滤波方法.首先,通过多尺度形态学开运算选择大量种子地面点;然后,基于种子地面点,使用布料模拟法由低至高逐层构建参考地形,以快速获取高分辨率参考地形;最后,基于点至参考地形的高差区分地面点和非地面点.利用国际摄影测量和遥感学会提供的数据集和参考方法,评估该方法性能.利用在中国、美国多个代表性林区的点云数据,评估该方法的可推广性.结果表明,该方法的Kappa系数和运行时间是83.72％和34.11 s,精度和效率较经典布料模拟滤波方法提高10.49％和52.17％.相比8种参考方法,该方法能够获得更高精度,并且具有稳定的可推广性.
Multi-resolution Hierarchical Cloth Simulation Filtering for Airborne LiDAR Data in Forested Areas
To obtain better filtering performance for complex forested sites,this study proposes a multi-resolution hierarchical cloth simulation-based filtering method for airborne light detection and ranging data.First,abundant ground seeds are selected using multi-scale morphological method,providing detailed basis for reference terrain initialization.Then,the reference terrain is constructed by hierarchical cloth simulation method for rapidly obtaining high-resolution reference terrain.Finally,ground points are extracted by comparing the height differences between points and reference terrain.We evaluate the performance of the proposed method using reference datasets and methods provided by the international society for photogrammetry and remote sensing.Meanwhile,the datasets of several representative forested areas in China and the United States are used to evaluate the generalizability of the proposed method.Results show that the Kappa coefficient and the runtime of the proposed method is 83.72％and 34.11 seconds,respectively,which is 10.49％and 52.17％higher than that of the classical cloth simulation filtering method.Compared with eight reference methods,the proposed method can obtain higher accuracy and generalization.
airborne light detection and ranging(LiDAR)forested areafilteringmorphologycloth simulation
万方数据
维普
