首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于最小负曲率点搜索的快速人体肢体提取算法
引用本文:曹丹华,汤金宽,吴裕斌.基于最小负曲率点搜索的快速人体肢体提取算法[J].华中科技大学学报(自然科学版),2007,35(5):16-18.
作者姓名:曹丹华  汤金宽  吴裕斌
作者单位:华中科技大学,光电子科学与工程学院,湖北,武汉,430074;华中科技大学,光电子科学与工程学院,湖北,武汉,430074;华中科技大学,光电子科学与工程学院,湖北,武汉,430074
摘    要:人体肢体分割涉及两类最小负曲率点,利用人们对人体肢体划分的主观意识,提出最小负曲率点分析法,搜索分割手臂和腿部的最小负曲率点.跟踪目标前景区域的边缘,得到前景区域的外轮廓.采用傅里叶算子对人体轮廓实现平滑处理,消除局部噪声对最小负曲率点的影响.采用移动向量法求取轮廓曲线的数字曲率.在划分出的各个肢体部分内部进行轴线搜索,确定肢体关节点,完成人体肢体的提取.实验结果表明:算法可有效搜索最小负曲率点,实现人体肢体分割.

关 键 词:图像分析  视频监控  肢体提取  最小负曲率点
文章编号:1671-4512(2007)05-0016-03
修稿时间:03 6 2006 12:00AM

Algorithm for fast extracting human limb contours using searching NCM points
Cao Danhua,Tang Jinkuan,Wu Yubin.Algorithm for fast extracting human limb contours using searching NCM points[J].JOURNAL OF HUAZHONG UNIVERSITY OF SCIENCE AND TECHNOLOGY.NATURE SCIENCE,2007,35(5):16-18.
Authors:Cao Danhua  Tang Jinkuan  Wu Yubin
Institution:School of Optoeletronics Science and Engineering, Huazhong University of Science and Technology, Wuhan 430074, China
Abstract:Human limb contour Extracting is related with two kinds of negative curvature minima(NCM).By introducing subjective judgment of the limb extracting, an algorithm of NCM is proposed to search the negative curvature minima on the limb contour. The profile of targets could be extracted by tracing their foreground.Fourier filter is implemented to smooth the profile and reduce the local noise effects on NCM.Besides,Moving vector analysis is introduced to compute the digital curvature of the profile.Also,further search of axis in local body profile is implemented to locate the joints and complete body extracting.The result of experiments showed that our algorithm provided distinct improvement in effective NCM search and human body extracting.
Keywords:image analysis  visual surveillance  limb extracting  negative curvature minima
本文献已被 CNKI 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号