多视点云配准是三维激光扫描数据处理过程中一项非常重要的任务,经典ICP算法是多视点云配准算法中一项重要方法。重点阐述了ICP算法的原理、解算步骤与推导过程,针对经典ICP算法不适用于大旋转角度的缺陷,进行了有效改进,在进行ICP算法解算前先进行一次粗配准,并对通过三维激光扫描采集的实测点云数据进行验证,实验表明该改进算法正确有效,适用于任意旋转角度的多视点云配准。
Multi-view point clouds registration is a critical task of three-dimensional laser scanning and the classical ICP algorithm is an important method of multi-view point cloud registration algorithm. The paper expounds the principle, calculation step and derivation process of ICP algorithm. In order to overcome the defects of large rotation angle of the classical ICP algorithm, coarse registration is done before the ICP algorithm calculation and real point cloud date acquired by 3D laser scanning is verified. The experimental results show that the improved algorithm is correct and effective, and can be used for multi-view point clouds registration with arbitrary rotation angle.