基于k-邻域密度的离散点云简化算法与实现  

Simplification and Implementation of Scattered Point Cloud Based on the Density in k-Nearest Neighbors

作　　者：车翔玖[1] 刘杨[1] 赵义武[2] 车娜[1] 高占恒[1] 

机构地区：[1]吉林大学计算机科学与技术学院,长春130012 [2]吉林大学数学研究所,长春130012

出　　处：《吉林大学学报（理学版）》2009年第5期994-998,共5页Journal of Jilin University:Science Edition

基　　金：国家自然科学基金(批准号:60773098;60673021);教育部博士学科点专项基金(批准号:20060183041)

摘　　要：提出一种基于k-邻域密度(即k-邻域中的点云密度)的离散点云简化算法,并给出了在三角网格重构中的实现.该方法不仅可以保证实物模型重建后的整体轮廓,而且在细节部分也较好地保持了局部形状特征.三角网格重构的实验结果表明,所给方法简单、高效,同时,在实物模型平滑处与曲率变化较大处均取得了理想效果.This paper presents a simplification method for scattered point cloud based on the density of the points in k-nearest neighbors as well as a concrete algorithm and the effect of its implementation on triangulation. This method ensures that reconstruction of triangular mesh maintains the figure contour of an object, but also does well on the details of a local part. The examples show that the method in this paper is easy to implement, and holds high efficiency. Meanwhile, this method holds satisfied effect on the smoothness part, even on the part where the curvature varies greatly.

关 键 词：k-邻域 离散点云 点云密度 点云简化 三角网格重构 

分 类 号：TP391.72[自动化与计算机技术—计算机应用技术] O241.5[自动化与计算机技术—计算机科学与技术]