CAJ | 학술논문

经典的渐进三角网滤波算法在LiDAR点云数据处理中应用十分广泛，但其滤波精度很大程度上取决于种子点选取的正确率。本文针对这一问题，在渐进三角网加密算法基础上提出了一种基于小格网高程、均方差和点密度统计数据选取种子点的迭代滤波算法。实验结果表明，本文的迭代滤波算法可有效避免低点等非地面点对种子点选取的干扰，且滤波结果生成的DEM精度较高，具有一定的实用性。
경전적점진삼각망려파산법재LiDAR점운수거처리중응용십분엄범，단기려파정도흔대정도상취결우충자점선취적정학솔。본문침대저일문제，재점진삼각망가밀산법기출상제출료일충기우소격망고정、균방차화점밀도통계수거선취충자점적질대려파산법。실험결과표명，본문적질대려파산법가유효피면저점등비지면점대충자점선취적간우，차려파결과생성적DEM정도교고，구유일정적실용성。
The classical adaptive TIN filtering algorithm is widely adopted in Data processing of LiDAR point clouds ,but the filter ac-curacy largely depends on the accuracy of seed point selection .Aiming to solve this problem , an iterative filtering algorithm of using statistical data of small grid ’ s elevation , Standard Deviation and point density choose seed points is stated in this article based on the classical adaptive TIN filtering algorithm .Experimental results show that the iterative filtering algorithm can effectively avoid the inter -ference of Non ground points such as low point to seed point selections .Also, the DEM that results from filter has a rather high preci-sion and practicality .