基于不同体型的三维模型自动分割方法研究

The study of 3D model automatic segmentation method based on different body dimensions

对不同体型的人体点云数据进行三维模型分割是三维人体测量的难点问题,提出了一种基于不同体型的人体三维模型自动分割方法。针对人体不同连接部位的形态特点差异,将近似凸包边最大距离分割法、轮廓部位曲线迭代拟合分割法和躯干切面特征点分割法相结合,实现人体表面曲率结构特征和截面轮廓特征融合的分割方法。分割结果表明,本方法具有自动、快速和易操作等特点,既保证了分割位置的准确性,也实现了人体模型分割的普适性要求。

3D model segmentation according to the point cloud data of different body dimensions is a difficult problem in3 D body measurement,a 3D model automatic segmentation method based on different body dimensions is presented in this study. A segmentation method is implemented by fusing the body surface curvature structural characteristic and cross section profile characteristic,which is based on the morphological characteristics difference of the different parts of human body. The segmentation method is combined with approximate convex-edge-maximum separation method,contour partial curve iteration fitting segmentation method, trunk segmentation feature point segmentation method. The segmentation result demonstrates that this method presents the characteristics of automatic,quickly and easy-operation,which can not only ensure the accuracy of the segmentation location but also achieve the universal requirements of human body model segmentation.