摘要： 研究激光扫描中的点云数据重构技术，提出一种基于规则点云数据的快速曲面重构方法。分析相邻扫描线之间数据点的相对位置关系，在三角剖分的基础上，设计改进的扫描线剖分算法，根据激光逐行扫描的特点，对点云数据进行不规则三角网格划分，利用几何关系进行配对构网，并在所建三角模型的基础上实现三角网格的局部优化和纹理映射，得到重建模型。实验结果表明，与传统Delaunay空间三角剖分算法相比，该算法可明显提高三角构网速度和质量，消除空洞，改善重建效果。

关键词: 激光扫描, 三角构网, 点云数据, 纹理映射, 曲面重建

Abstract: This paper researches the technonlogy of point cloud data reconstruction in laser scanning, and proposes a fast method of curved surface reconstruction based on point cloud. It analyses the relative position of the relationship of data points between adjacent scanning lines, and improves scanning lines subdivision algorithm based on triangular subdivision, according to the characteristics of the progressive scanning laser scanning. The point cloud data is irregular triangular mesh divided. It uses geometrical relationship to make triangular network construction. At the same time, on the basis of the triangle model, the triangular mesh is optimized locally, to make texture mapping, and to be rebuild in the end. Experimental results show that this algorithm can effectively improve the speed and quality of triangular network construction, eliminate voids and improve the reconstruction results.

Key words: laser scanning, triangular network construction, point cloud data, texture mapping, curved surface reconstruction

中图分类号: 

• TN911.73