基于粒子群优化算法的散乱点云数据配准  

Registration of Scattered Cloud Point Data Based on PSO

作　　者：段德全[1] 李俊芬[2] 申培萍[2] 

机构地区：[1]河南师范大学计算机与信息技术学院,河南新乡453007 [2]河南师范大学数学与信息科学学院,河南新乡453007

出　　处：《广西师范大学学报（自然科学版）》2008年第3期226-229,共4页Journal of Guangxi Normal University:Natural Science Edition

基　　金：国家自然科学基金资助项目(10671057)

摘　　要：针对目前散乱点云数据配准算法在精度、速度和优化等方面存在的问题,提出一种基于粒子群优化算法的点云数据配准算法。该算法首先根据数据点之间曲率的相似度函数,采用粒子群优化算法在两组点云数据中搜索可以匹配的点对集合,然后用最近点迭代算法进行二次配准,实现了两组散乱点云数据的精确配准。对比实验表明,该算法配准速度快,效果好。A point cloud registration algorithm based on the particle swarm optimization algorithm is proposed, for the deficiencies of precision, speed and optimization of present registration schemes. All the pair wise points are first located whose curvatures were sufficiently similar based on the curvature similarity function, by using the particle swarm optimization algorithm. Then, a perfect registration is deduced by using the iterative closest point algorithm. Contrast experiments show the high speed and robust performance of the proposed algorithm.

关 键 词：粒子群优化算法 散乱点云 曲率 配准 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]