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

用于步态识别的行人轮廓提取
引用本文:佘艳,赵玥,贲晛烨,李阳.用于步态识别的行人轮廓提取[J].应用科技,2012(2):19-25.
作者姓名:佘艳  赵玥  贲晛烨  李阳
作者单位:1. 顺德职业技术学院电子与信息工程系,广东顺德528333
2. 哈尔滨工程大学自动化学院,黑龙江哈尔滨150001
3. 哈尔滨工业大学交通科学与工程学院,黑龙江哈尔滨150090
4. 一汽轿车股份有限公司,吉林长春130011
基金项目:中国博士后科学基金面上项目(20110491807)
摘    要:采用针对静态背景下的基于Surendra背景更新算法的背景减除法对运动人体进行检测.为视频场景建立自适应的背景模型,通过原始图像和背景模型差分获得前景图像,再对检测出来的图像进行了二值化、数学形态学分析、连通分析、尺度归一等一系列图像预处理工作,为跟踪与识别奠定了基础.重点讨论了二值化自适应阈值选择的多种方法,总结出Kapur熵阈值选取法的优越性.

关 键 词:步态识别  轮廓提取  背景更新  二值化

Silhouette extraction of a moving body for gait recognition
SHE Yan,ZHAO Yue,BEN Xianye,LI Yang.Silhouette extraction of a moving body for gait recognition[J].Applied Science and Technology,2012(2):19-25.
Authors:SHE Yan  ZHAO Yue  BEN Xianye  LI Yang
Institution:1.Department of Electronic and Information Engineering,Shunde Polytechnic,Foshan 528333,China 2.College of Automation,Harbin Engineering University,Harbin 150001,China 3.School of Transportation Science and Engineering,Harbin Institute of Technology,Harbin 150090,China 4.Faw Car Limited Company,Changchun 130011,China
Abstract:This paper presents a background subtraction method for detecting a moving body under the static background based on Surendra’s background updating algorithm.An adaptive model of background was established for the video scene,then a foreground image was obtained by the differential method between the original image and the background model.And then the image pre-processing was carried out,including binarization,mathematical morphology analysis,connectivity analysis,scale normalization,etc.,which lays a foundation for further tracking and identifying.The emphasis was discussing a discussion about the ways to choose the adaptive binary threshold.A summary of the advantages of using Kapur entropy threshold selection method was given.
Keywords:gait recognition  silhouette extraction  background updating  binarization
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