三维激光点云配准算法研究
Research on 3D Laser Point Cloud Registration Algorithm

为提高点云配准的效率与精度，弥补传统点云配准算法中的不足，本文尝试将主方向贴合算法与改进的最近点迭代(Improved Iterative Closest Point, IICP)算法作为组合点云配准算法；该组合算法充分利用主方向贴合算法在粗配准中的优势，并结合IICP算法作为精配准的方法。以某三维激光扫描建模工程为案例，依靠Matlab软件编程实现了本文所提及的配准算法。实验结果表明，改进的ICP算法较传统ICP算法配准效率与精度均有提高。充分验证了主方向贴合算法与IICP点云算法在点云三维模型构建方面的有效性。
To improve the efficiency and accuracy of point cloud registration and make up for the shortcom-ings of traditional point cloud registration algorithms, this paper attempts to use the principal di-rection fitting algorithm and the improved Iterative Closest Point (IICP) algorithm as a combined point cloud registration algorithm; the combined algorithm fully utilizes the advantages of the principal direction fitting algorithm in rough registration, and combines IICP algorithm as a fine registration method. Taking a 3D laser scanning modeling project as an example, the registration algorithm mentioned in this article was implemented by programming with Matlab software. The experimental results show that the improved ICP algorithm has improved registration efficiency and accuracy compared to the traditional ICP algorithm. The validity of the principal direction fit-ting algorithm and the IICP point cloud algorithm in constructing 3D model of point clouds is fully verified.