逆向工程中一种鲁棒的平移对称提取算法

Robust Algorithm for Translation Symmetry Extraction in Reverse Engineering

为了解决逆向工程中平移对称提取的问题,提出了一种鲁棒的提取算法.首先利用特征线将模型分割成体素的集合,然后计算各个体素的形状分布曲线,通过形状分布曲线匹配的方式获取了几何形状相似的体素,并利用最小包围盒从中提取大小相似的体素,以构成对称体素的集合.利用鲁棒的PCA方法估算对称体素的初步对称信息,在此基础上利用迭代最近点算法对提取的平移对称进行优化.采用一个机械零件的三角网格模型进行实验的结果表明,体素分割结果对应构成模型的各个特征,具有明显的工程语义,对称提取结果的误差较小并具有较强的鲁棒性.

To solve the problem of the translation symmetry extraction in reverse engineering,a robust algorithm is presented.The model is segmented into the set of voxels based on its feature lines,then for each voxel,the shape distribution histogram is calculated with D2 shape distribution function.Comparing each pair of shape distribution histogram by curve matching,the voxels with similar geometric shape are acquired and the set of symmetric voxels are constructed by ones having similar minimum bounding box.The initial symmetric information is estimated by the robust principle component analysis and optimized with the algorithm of iterative closest point.An experiment with triangular mesh model is conducted,and the result shows that the voxels of the segmentation corresponding to the features composed of the model demonstrate the obvious engineering semantics and the symmetrically extracted information gets accurate and robust.