基于LiDAR点云数据分布差异的植被三维实景建模算法

Three-Dimensional Modelling Algorithm of Real Scene Vegetation Based On Lidar Point Distributing Difference

传统的植被三维建模算法难以提取出准确的树木生长和几何形态参数。本文提出了基于LiDAR点云数据分布差异的植被三维实景建模算法,首先运用激光点云数据和高光谱信息分离枝干和叶片,其次分别对植被主干和分支的分布差异采用不同算法建模,构建出能够反映树木真实几何结构参数的三维模型,最后进行对比实验。实验结果表明,与已有算法相比,本文算法能更好地保持树木主干的外形轮廓,且分枝模型之间的空间拓扑关系基本与实际相符,数据量更少,处理效率更高。

Traditional vegetation 3D modeling algorithms are difficult to extract accurate tree growth and geometric morphological parameters. In this paper, we propose a 3D realistic modeling algorithm for vegetation based on the distribution differences of LiDAR point cloud data. The experimental results show that compared with the existing algorithms, the algorithm in this paper can better maintain the contour of the tree trunk, and the spatial topological relationship between the branch models is basically consistent with the actual one, and the data volume is smaller and the processing efficiency is higher.