机载激光雷达点云辅助的镶嵌线自动选择

本文提出了一种基于激光雷达点云辅助的正射影像镶嵌线自动选取方法。首先,利用配准后的激光雷达点云生成数字表面模型和数字高程模型,以重叠区域影像之间的相似度和梯度差异构建代价矩阵;然后,将数字表面模型与数字高程模型相减得到差值图表示地物真实的高程信息,利用自适应阈值分割将差值图划分为不同的区域,并赋予不同的惩罚系数;最后,以代价最小为标准,基于Dijkstra算法得到最终的镶嵌线。本文方法生成的镶嵌线能成功地绕开大多数障碍地物,能够满足影像拼接应用的需求。

Automatic seamline selection assisted by airborne LiDAR

An automatic orthophoto seamline selection method assisted by LiDAR is proposed in this paper. The LiDAR point cloud is used to generate DSM and DEM, the difference between which is used to obtain the difference map to represent the true height of the object. The threshold is adaptively set to divide the difference map into different regions, and different weights are given to guide the automatic selection of the seamline based on Dijkstra algorithm. The experimental results demonstrate that the proposed method is robust to the selection of parameters and the final seamline can successfully bypass the obstacles.