特征保持的散乱点集光顺算法  

Feature-preserving smoothing algorithm for scattered point data

作　　者：夏仁波[1] 刘伟军[1] 王华兵[1] 

机构地区：[1]中国科学院沈阳自动化所先进制造技术实验室,沈阳110016

出　　处：《仪器仪表学报》2006年第z1期7-9,共3页Chinese Journal of Scientific Instrument

基　　金：国家"863"高科技(2002AA420060-1)资助

摘　　要：提出了一种特征保持的散乱点集光顺算法。首先,搜索点的k最近邻域,计算出点的Delaunay邻域。然后通过考查点及其二阶邻域之间的几何关系,设计出一种带有抑制函数的双边滤波器,用于噪声点集模型的光顺。实验结果显示,本文的光顺算法不但能够有效地去除噪声,抑制体积收缩,并对特征的保持效果也较为理想。A feature-preserving smoothing algorithm for scattered point data is proposed.First,the k nearest neighbors of a point are checked to compute the Delaunay neighbors of the point.Then,a bilateral filtering with prohibitive function is designed to smooth the noisy point data based on the geometric analysis of point's relation with the specific local neighbors.The experimental results show that new smoothing algorithm can filter out the noise in the point data,preserve the feature well,prohibit the volume from shrinking,and is more efficient than previous methods.

关 键 词：散乱点 噪声 光顺 特征保持 

分 类 号：TH7-55[机械工程—仪器科学与技术]