足部模型点云配准方法研究

Research on Point Cloud Registration Method of Foot Model

由于人体足部呈长条状且不规则的几何形状,使得用三维扫描设备进行点云数据采集时,会出现点云数据重叠度较低的情况,配准难度较高。针对人体足部的外表结构不规则几何特性,提出了一种先粗配准再细配准方法对足部模型点云进行配准。先对采集到的足部模型点云数据进行去噪、滤波等预处理,再在足部模型下设置一立方体底座,利用立方体底座法向量规则均匀的特性来辅助粗配准工作;再采用点云分割方法去掉立方体底座,最后利用ICP最近点迭代对足部模型进行细匹配,从而实现对足部模型的点云配准。实验结果表明:提出的配准方法与传统的ICP方法相比,不但可以实现对足部模型点云的配准,且在配准速度上提高了54%,配准精度上提高了42%。

Because the foot of human body is long and irregular geometry, when using 3 D scanning equipment to collect point cloud data, there will be low overlap of point cloud data, and registration is difficult. According to the irregular geometry of human foot, a method of rough registration and fine registration was proposed to register the point cloud of foot model. Firstly, preprocess the collected point cloud data of foot model, such as preprcess and filtering, then set a cube base under the foot model, use the regular and uniform characteristics of the normal vector of cube base to assist the coarse registration;then use the point cloud segmentation method to remove the cube base, and finally use the ICP nearest point iteration to fine match the foot model, so as to realize the fine matching of the foot model Point cloud registration. The experimental results show that the proposed registration method can not only achieve the registration of point cloud of foot model, but also improve the registration speed by 54% and the registration accuracy by 42% compared with the traditional ICP method.