基于数据相关性的STL曲面网格快速重建算法

Fast Reconstruction of STL Surface Meshes Based on Data Correlation

通过分析大量STL文件中的三角形单元数据,发现文件中顺序相邻的2个独立的三角形网格单元至少共一个顶点的概率大约是0.84～0.99,共2个顶点的概率大约是0.67～0.75,表明相邻网格单元数据存在强相关性以及大量的冗余信息.利用这种数据相关性,从概率的角度给出了一条检查冗余点的有效途径,进而得到一种快速的STL三角形曲面网格重建算法.新的曲面网格数据文件存储容量为原来的25%左右,有效地去除了冗余数据.实验结果表明了该算法的高效性及鲁棒性.

By analyzing data of triangular elements in a large number of STL files, we found that the probability that two orderly adjacent triangular elements share at least one vertex is about 0.84 - 0. 99, and the probability of sharing two vertices is about 0. 67 - 0. 75. This indicates that strong correlation exists in adjacent triangle elements, and there is a great deal of redundant data in files. By using the data correlation, this paper presents an effective approach to check redundant points from the perspective of probability and a fast surface reconstruction algorithm of the STL triangular meshes. Experimental results show that, because the redundant data is eliminated effectively, the new storage capacity of surface mesh data file is 25 percent of that of the original file.