%0 Journal Article
%T 基于改进ISS特征点与人工蜂群算法的点云拼接方法<br>Point Clouds Registration Algorithm Based on Improved ISS Feature Points and Artificial Bee Colony Algorithm
%A 葛宝臻
%A 周天宇
%A 陈雷
%A 田庆国
%J 天津大学学报（自然科学与工程技术版）
%D 2016
%R 10.11784/tdxbz201509022
%X 传统ICP算法在进行点云拼接时易陷入局部最优，利用群智能的优化方法可以解决这一问题，但同时会带来计算量较大的问题．为此，本文首先提出了一种新的基于人工蜂群(ABC)优化的点云拼接方法，通过引入邻域半径约束的改进固有形状特征点提取方法对初始模型进行简化，然后采用人工蜂群算法对简化后模型间对应点的欧几里德中值距离进行优化求解，得到空间变换矩阵T的参数，将变换矩阵作用于原始模型，从而完成对点云的高效拼接．通过对不同初始位置的理想点云库模型以及实际扫描的带有噪声的人体点云模型进行拼接实验，结果表明本文算法不仅对于理想模型具有很高的精度，对于实际获得的点云模型也有很好的抗噪性，而且计算效率比采用全部点计算的群智能算法提高了6倍．<br>Traditional ICP algorithm tends to fall into local optima in point clouds registration，while optimization algorithms based on swarm intelligence can solve this problem．However，the swarm intelligence algorithms involve a large amount of calculation．In order to solve the above problems，we propose a novel point clouds registration algorithm based on artificial bee colony(ABC) algorithm．First of all，we use an improved ISS feature point extraction algo-rithm adopting a neighbour points radius constraint strategy to simplify the models．Then we use the ABC algorithm to obtain the parameters of T to complete the efficient registration．Through the registration results of ideal models with different initial position and real human model with noise used in this paper，it is found that the performance of our algorithm acquires accurate results in deal models and keeps robust in real models．The computational efficiency is 6 times improved compared with that of traditional swarm intelligence optimization algorithm
%K 点云拼接
%K 特征点提取
%K 人工蜂群算法
%K 全局收敛
%K 运算效率&lt
%K br&gt
%K point clouds registration
%K feature points extraction
%K ABC algorithm
%K global convergence
%K computational efficiency
%U http://journals.tju.edu.cn/zrb/oa/darticle.aspx?type=view&id=201612012