基于点云数据的3D人体模型骨骼提取  

Algorithm for Human Skeleton and Joints Based on Scanned Human Body Cloud Data

作　　者：黄三艳[1] 李雪飞[1] 

机构地区：[1]北京服装学院,北京100029

出　　处：《北京服装学院学报（自然科学版）》2014年第2期16-20,共5页Journal of Beijing Institute of Fashion Technology：Natural Science Edition

摘　　要：骨架提取是构建真实人体模型动画的前提和重要条件.本文以基于测地距离的Morse函数和Dijkstra算法计算出末端特征点信息.通过测地距离等值线将模型进行分割,计算每一层的测地距离等值线,获取模型的也就是水平集曲线(level-set-curve).以末端特征点做为起点,拟合相邻水平集曲线的重心形成中心线(模型的骨架).采取截面似圆性标准获取模型关节中心所在的截面,求取该截面的重心,从而获取模型的关节中心.Skeleton and joints is necessary for creating real human body animation. So we have to extract the skeleton and joints of scanned human body. Bases on Morse function, the paper used Di- jkstra Algorithm to calculate the end feature points of human body. We also used Morse function starting from head future point to cut the human body to many pieces in order to get the level-set- curves. By calculating the center point of gravity of each level-set-curve, and joining all the level- set-curve＇s center points with the end feature points of human body, we can get the center line called human skeleton. As the shape of each level-set-curve＇ s similarity of ring is different based on the distance of joint, so we can use this method to make a calculating which level-set-curve the joint center belongs to. By calculating the center point of gravity of joint＇ s level-set-curve, we can get the location of joint center.

关 键 词：点云数据 末端特征点 水平集曲线 中心线 关节中心 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]