一种顾及地形复杂度的LiDAR点云多尺度滤波方法

发布时间：2018-01-24 10:41

本文关键词： 机载LiDAR DEM 点云滤波自适应阈值薄板样条　出处：《测绘科学》2016年11期 　论文类型：期刊论文

【摘要】：针对复杂地形区域的机载LiDAR数据滤波方法中自适应阈值设置问题,根据地形多尺度效应,提出一种自适应阈值的机载LiDAR点云多尺度滤波方法。该方法采用影像金字塔策略按分辨率从高至低逐级构建LiDAR点云分层格网,滤波过程则从最大尺度格网(顶层格网,最低分辨率)开始,采用局部统计分析的方法自适应地确定高差阈值,同时结合薄板样条内插出下层各格网控制点的高程值,直至最底层格网完成原始激光点云滤波。通过我国某山区城市复杂地形的LiDAR数据实验表明顾及地形复杂度的LiDAR点云多尺度滤波方法能够快速有效地提取高精度DEM,能够满足实际生产需求。
[Abstract]:Aiming at the problem of adaptive threshold setting in airborne LiDAR data filtering in complex terrain regions, the multi-scale effect of terrain is considered. An adaptive threshold multi-scale filtering method for airborne LiDAR point clouds is proposed, in which the hierarchical grid of LiDAR point clouds is constructed by using image pyramid strategy from high to low resolution. The filtering process starts with the maximum scale grid (the top grid, the lowest resolution), and adaptively determines the threshold of height difference by using the method of local statistical analysis. At the same time, the height of each grid control point in the lower layer is interpolated with the thin plate spline. The LiDAR data of complex terrain in a mountainous city in China show that the multiscale LiDAR filtering method with terrain complexity can be quickly obtained by using the original laser point cloud filter until the bottom grid is completed. High precision DEM was extracted effectively. Able to meet actual production requirements.
【作者单位】：湖南科技大学地理空间信息技术国家地方联合工程实验室;武汉大学测绘遥感信息工程国家重点实验室;
【基金】：国家自然科学基金项目(41401497) 湖南省科技计划项目(2015GK3027)
【分类号】：P237
【正文快照】： 0引言机载激光雷达(light detection and ranging,LiDAR),是一种新兴的主动式遥感手段,通过它可以快速、高效、精准地获取地表及其地物目标表面离散采样点的三维坐标信息。由于LiDAR点云包含了目标区域内的所有地表、地物三维信息,如地形表面、建筑物、桥梁、植被、电力线、汽

【相似文献】