基于相似性的点模型简化算法

Similarity-based simplification of point-sampled surfaces

为了获得高质量的简化点模型,提出了一种基于相似性的曲率自适应点模型简化算法,相似性包括强特征边性和表面区域几何特征相似性2个方面.利用法向张量投票方法,计算采样点的特征边性,由此将点模型分为强边性和非强边性2部分;基于Mean Shift聚类法,对非强边性部分进行表面区域几何特征相似性聚类;对强边性部分和各类簇重采样,实现曲率自适应的简化,并通过移动最小二乘曲面,评估简化曲面的误差.实验结果表明,该算法有效地保持了特征边界部分和曲面的细节,且能够生成高质量的简化点集曲面.

In order to achieve a high-quality simplified model, an adaptive curvature simplification algorithm for point-sampled surfaces was presented based on similarity including strong feature-edge intensity and surface feature anisotropy. Using the normal tensor voting, the feature-edge intensity of sample points was evaluated, by which the point-sampled surfaces were segmented into two parts, one for the strong featureedge intensity and another for the nonstrong feature-edge intensity. Base on Mean Shift clustering, the second part was clustered into clusters according to the surface-features similarity. The first part and all the clusters were resampled in combination with the surface variation and sampling-density control so as to generate the simplified point set. In addition, the quality of the simplified point set surfaces was evaluated using the error measurement method based on the moving least squares surfaces. Experimental results showed that this algorithm not only can effectively preserve the feature edges and the surface details, but also can achieve the simplified point set generating high-quality surface approximations to the origin point set surfaces.