基于约束树形图结构外观模型的人体姿态估计  

Human Pose Estimation Based on Appearance Model for Constraint Tree Pictorial Structure

作　　者：王浩[1] 刘则芬 方宝富[1] 陈金金[1] 

机构地区：[1]合肥工业大学计算机与信息学院,合肥230009

出　　处：《计算机科学》2014年第3期76-79,共4页Computer Science

基　　金：国家自然科学基金(61070131;61175051;61175033;61075076);国家高技术研究发展计划(863计划)(2012AA011005);安徽省自然科学基金(1308085QF108)资助

摘　　要：人体姿态估计是计算机视觉领域中的一个关键问题,可用于行人检测、人体活动分析、人机交互以及视频监控等方面。针对基于树形图形结构模型的人体姿态估计算法中人体部件外观模型容易受到背景干扰的问题,提出了基于先验分割和外观转换的外观模型的人体姿态估计算法以改进人体的外观模型。根据PS模型,使用人体检测器和前景加亮进行预处理,确定人体的大致位置和大小,同时移除背景上的杂斑,基于先验分割和外观转换机制来估计人体部件的外观模型。实验表明在不同的图像数据库中,该算法在使用人体检测器和前景加亮算法减少部件搜索空间的同时,提高了人体姿态估计的准确性。Human pose estimation is one of the hot topic in the field of computer vision,and can be used for pedestrian detection,human activity analysis,human-computer interaction and video surveillance.Aiming at the problem that the human body appearance model of human parts is vulnerable to background interference in the body pose estimation algorithm of the tree pictorial structure model,we put forward body pose estimation algorithm of appearance model which is based on priori segmentation and the appearance transfer mechanism in order to improve the human body appearance model.According to PS model,using the human body detector and foreground highlighting by preprocessing,it can find an approximate position and size of the human body,meanwhile remove the background clutter.We can estimate the appearance model of the human body parts based on a priori segmentation and appearance transformation mechanism.Experiments show that using the algorithm of the human body detector and foreground highlighting can not only reduce the search space of the components,but also improve the accuracy of the body pose estimation.

关 键 词：图像处理 人体姿态估计 外观模型 图形结构 先验分割 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]