基于粒子群优化的点云场景拼接算法

多视点三维点点云场景拼接是解决激光三维主动成像目标自遮挡或被遮挡情况下目标数据不完全问题的一种有效方法，它将直接影响到后续的目标检测与识别处理。本文提出了一种基于粒子群优化（PSO）的点云拼接算法，该方法通过投影分布熵对待拼接场景姿态进行估计，由此获取场景拼接的初值，然后利用PSO实现多视点场景的精确拼接。仿真实验结果表明，本文方法是一种有效可行的方法。

Particle swarm optimization based point cloud scene mosaic algorithm

Multi-view point cloud scenes mosaic is an effective method to solving the incomplete object data problem while self-occlusion and occlusion happened in ladar 3D imaging process. Mosaic method directly affects the object detection and recognition. In this paper, a particle swarm optimization(PSO)based mosaic algorithm is proposed. In the paper, we use projective distribution entropy to estimate the transformation between point cloud scenes. Based on this initial value, the PSO is used for precise mosaic. Experimental results demonstrate the effectiveness an feasibility of the proposed algorithm.