基于局部特征的四点一致集配准算法研究

针对四点一致集点云配准算法的速度慢等问题,本文提出一种局部特征四点一致集算法。利用体素滤波器进行下采样,选取局部特征点;原点云和特征点进行快速点特征直方图对比,得到相似的特征;最后在特征点集上采用四点一致集算法,并与4PCS算法进行比较。选取3组点云数据进行分析,实验结果表明,基于特征点集的4PCS算法要优于4PCS算法,在运算速度和精度上有了提高。

Research on Four-point Uniform Set Registration Algorithm Based on Local Features

Aiming at the problem of slow speed of four-point uniform set point cloud registration algorithm.In this paper,a four-point uniform set algorithm for local features is proposed,which USES a voxel filter to carry out down-sampling and select local feature points.Fast histogram comparison of feature points and origin clouds was performed.Finally,a four-point consistent set algorithm is implemented on the feature point set and compared with the 4PCS algorithm.Three sets of point cloud data were selected for analysis,and the experimental results showed that the 4PCS algorithm based on feature point set was superior to the 4PCS algorithm,with improved operation speed and accuracy.