采用轮廓识别法实现零件曲面重构

Research on Surface Reconstruction Based on Linear Surface

线性曲面反求算法的目的是根据原始数据点云反求得到能够正确反映原零件拓扑结构的线性曲面模型。该反求算法的核心思想是构造一个空间函数 ,该函数近似表示三维空间上任一点到所求曲面的距离 ,然后通过轮廓识别算法 ,可以由这个函数提取反求曲面的一个线性近似。本文所阐述的线性曲面反求算法结合网格优化算法可由原始数据点云 ,反求得到一个精确而简洁的具有任意拓扑类型的零件线性曲面。

Through the surface reconstruction algorithm described in this paper, an accurate and concise linear surface of a part in arbitrary topology structure can be reconstructed from original data points. The goal of linear surface reconstruction is to reconstruct linear surface that can correctly represent topology structure of original object from data points. The key of the algorithm is to construct a space function that represents the distance from an arbitrary point in the space to the surface that we want to reconstruct. By using contour tracing algorithm, an iso-surface whose distance to the surface that we want to reconstruct is zero can be extracted by the function. The iso-surface is a linear approximation of the surface that we want to reconstruct. The algorithm is proved by actual application to be effective.