点云配准在三维人体与服装配准中的应用

Application of Point Cloud Registration in 3D Human Body and Clothing Registration

点云配准在逆向工程、计算机视觉和文物数字化等领域中应用广泛。针对传统ICP算法收敛速度慢且在两点云集初始位姿较大时易陷入局部极值的问题,提出一种基于ICP算法的改进点云配准算法,并将其应用于三维人体模型与服装模型的配准中。算法利用基于特征点的采样一致性初始配准算法(SAC-IA)实现两点云的初始配准,使两点云集有相对较好的初始位姿,同时利用ICP算法实现人体模型和服装模型之间的精确匹配。实验结果表明算法具有较好的配准精度和较快的配准速度。

Point cloud registration is widely used in reverse engineering,computer vision and cultural relic digitization.In view of the problems that the convergence speed of traditional ICP algorithm is slow and it is easy to fall into local extremum when the initial pose of two-point cluster is large,an improved point cloud registration algorithm based on ICP algorithm is proposed and applied to the registration of 3D human body and clothing model.The algorithm uses the sampling consistency initial registration algorithm based on feature points to realize the initial registration of the two-point cloud,so that the two-point cluster has a relatively good initial pose,as well as the ICP algorithm is used to realize the accurate matching between the human body and the clothing model.Experimental results show that the algorithm has better registration accuracy and faster registration speed.