切片厚度加权的二次误差测度网格简化算法

Mesh Simplification with Slice Thickness-Weighted Quadric Error Metric

针对目前三角网格简化效率低、模型表面细微特征丢失的现象,基于模型切片处理算法对经典二次误差测度算法进行改进.该算法采用半边结构来存储模型数据,应用代价最小的边折叠二次误差测度算法,分析点到相邻平面的距离,同时引入切片厚度加权因子来减少阶梯面的形成,从而保证了简化后模型表面的细微特征.实验结果表明,对于原始三角面片数超过6万的,该算法相较于经典的二次误差测度算法效率提高了9.2%,对于原始三角面片数不足1万的,该算法相较于经典的二次误差测度算法提高了7.1%,模型经大规模简化后表面细微特征得到了很好的保留.

Aiming at the low efficiency of triangular mesh simplification and the loss of subtle features on the model surface, the classical quadratic error metric algorithm was improved based on the model slice processing algorithm. The algorithm used the half-edge structure to store the model data, and used the least-cost edge collapse quadric error metric algorithm to analyze the distance from the point to the adja- cent plane, and the slice thickness weight was used to reduce the step surface, thus ensuring the subtle features of the simplified model surface. The experimental results show that the proposed algorithm is 9. 2 % higher than the classical quadratic error metric algorithm for the number of more than 60000 original triangular faces, and the proposed algorithm is 7. 1% higher than the classical quadratic error metric algorithm for the less than 10000 original triangular faces, and the fine features of the model are well preserved after large scale simplification.