中文摘要：

针对大规模建筑物点云数据采用CPD（coherent point drift）算法进行配准时,计算复杂度增大的问题,提出了一种基于建筑物点云特征点简化数据的快速配准ISS-CPD算法。该配准算法采用ISS（intrinsic shape signature）算法求得建筑物点云的特征点,可减少建筑物点云的数据量规模,再对所提取的不同视角下建筑物点云的特征点用CPD算法进行配准。实验结果表明,改进的配准算法提高了建筑物点云的配准效率。

英文摘要：

To counter the problem of the increasing computational complexity when using CPD（ Coherent Point Drift） algorithm for registration to the large-scale building point cloud data,this paper proposes a fast registration algorithm based on feature points to simplify the building point cloud. This point cloud registration algorithm is obtained by using the algorithm of ISS to extract building feature points to reduce the scale of building point cloud data,and use CPD algorithm to register the feature points of the multi-view building point cloud. The experimental results show that the improved registration algorithm is simple,effective,stable and reliable,which can greatly improve the registration efficiency of the building point cloud using CPD algorithm.