一种改进的基于TIN渐次加密的LiDAR点云滤波算法

在深入研究原始不规则三角网(TriangulatedIrregularNetwork,简称TIN)算法的基础上,提出了一种改进的TIN算法。该改进算法通过构建网格筛选地面种子点来提高初始面的可信度。在向上加密过程的四面体顶点阈值选取上,采用了基于局部地形坡度的自适应阈值,克服了原始算法预先设定单一阈值的局限性。实验结果表明,该改进算法可以有效降低三类误差,其中,I类误差降幅达50%,II类误差达30%,总误差达40%。

A Modified TIN-Based LiDAR Data Encryption Gradually Filtering Algorithm

This paper proposed a modified filtering method based on in-depth study of the original Triangulated Irregular Network (TIN) algorithm.It improves the credibility of the initial surface by constructing the grid to generate the ground seed points.When selecting the tetrahedron vertices threshold in the upward encryption process,it used adaptive threshold based on the local terrain gradient to overcome the limitation of single threshold in original algorithm.Experiments show that the modified algorithm can effectively reduce the three types of errors,including type Ⅰ error reduction of 50％;type Ⅱ error of 30％;the total error of 40％.