正在加载图片...
第8卷第5期 智能系统学报 VoL.8 No.5 2013年10月 CAAI Transactions on Intelligent Systems 0ct.2013 D0:10.3969/j.issn.1673-4785.201110007 网络出版地址:http://www.cnki.net/kcms/detail/23.1538.TP.20130929.1049.003.html 线性插值框架下矩阵步态识别的性能分析 贲晛烨,张鹏,潘婷婷,王科俊 (1.山东大学信息科学与工程学院,山东济南250100:2.哈尔滨工程大学自动化学院,黑龙江哈尔滨150001) 摘要:针对现有的步态周期检测方法检测效果不佳以及行走速度变化对步态识别性能有很大影响的问题,提出的 基于矩的步态周期检测方法中,Zernike矩需要人体居中、尺度归一的前期预处理过程,而伪Zernike矩具有能描述运 动图像的特点,它可以避免人体居中、尺度归一等处理,以便直接测试步态的周期性.根据行走时的两帧之间的特征 取决于前一帧和后一帧的特征,提出了基于线性插值的矩阵步态识别算法框架,并且将投影特征、Hough变换特征、 Trace变换特征和Fan-Beam映射特征应用在CASIA(B)步态库上,验证了框架的有效性,为解决步态识别问题带来 新的方法与思路这种基于线性插值的矩阵步态识别特征本质上是一种权值不同的能量形式. 关键词:步态识别:矩阵步态识别:线性插值框架;步态周期检测;Zernike矩;伪Zernike矩 中图分类号:TP391.41文献标志码:A文章编号:1673-4785(2013)05-0415-11 中文引用格式:贲晛烨,张鹏,潘婷婷,等.线性插值框架下矩阵步态识别的性能分析[J】.智能系统学报,2013,8(5):415425. 英文引用格式:BEN Xianye,ZHANG Peng,PAN Tingting,etal.Performance analysis of matrix gait recognition under linear in- terpolation framework[J].CAAI Transactions on Intelligent Systems,2013,8(5):415-425. Performance analysis of matrix gait recognition under linear interpolation framework BEN Xianye',ZHANG Peng',PAN Tingting',WANG Kejun2 (1.School of Information Science and Engineering,Shandong University,Ji'nan 250100,China;2.College of Automation,Harbin Engineering University,Harbin 150001,China) Abstract:The existing gait period detection methods are not ideal and the performance of gait recognition is signifi- cantly influenced by walking speed.Several novel gait period detection methods based on moments are proposed in this paper.The Zernike moment requires preprocessing including the assurance that the image of the human body is proportioned normally and is centered properly;the pseudo-Zernike moment may directly describe the motion im- age,and it may avoid the need for such processing of making the image of the human body centered and sized nor- mally,so as to directly detect gait periodicity.As the features of one frame are only decided by those of the prior and the rear frames in walking,a framework for a matrix gait recognition algorithm based on linear interpolation is proposed.Subsequently,the projection features,Hough transform feature,Trace transform feature and Fan-Beam mapping feature are applied to the CASIA(B)gait database to prove the validity of the gait recognition framework. This brings new methods and understanding for solving gait recognition problems.This matrix gait recognition feature based on linear interpolation is essentially an energy form with different weighted values. Keywords:gait recognition;matrix gait recognition;linear interpolation framework;gait period detection;Zernike moment;pseudo-Zernike moment 步态识别是在人没有觉察的远距离情况下,通 特征具有稳定性、普遍性、可采集性和惟一性,因此 过捕捉动态与静态信息来进行身份识别.由于步态 可用于不同场合下的身份识别与认证).Nixon等] 收稿日期:2011-10-19.网络出版日期:2013-09-29 总结了南安普敦大学2002年之前研发的步态识别 基金项目:国家自然科学基金资助项目(61201370):高等学校博士学 技术,同时预言了步态应用于生物特征识别的潜能, 科点专项科研基金资助项目(20120131120030):中国博土 后科学基金面上项目(2013M530321):山东省博土后创新 2005年,Boulgouris等I)首先将步态识别与其他生 项目专项资金项目(201303100):山东大学自主创新基金 物特征识别做了比较,提出步态识别可以作为多生 资助项目(2012GN043,2012DX007). 通信作者:贲現烨.E-mail:benxianyeye@l63.com 物特征识别技术的一部分,接着给出了步态识别系第 8 卷第 5 期 智 能 系 统 学 报 Vol.8 №.5 2013 年 10 月 CAAI Transactions on Intelligent Systems Oct. 2013 DOI:10.3969 / j.issn.1673 ̄4785.201110007 网络出版地址:http: / / www.cnki.net / kcms/ detail / 23.1538.TP.20130929.1049.003.html 线性插值框架下矩阵步态识别的性能分析 贲晛烨1 ꎬ张鹏1 ꎬ潘婷婷1 ꎬ王科俊2 (1.山东大学 信息科学与工程学院ꎬ山东 济南 250100ꎻ 2.哈尔滨工程大学 自动化学院ꎬ黑龙江 哈尔滨 150001) 摘 要:针对现有的步态周期检测方法检测效果不佳以及行走速度变化对步态识别性能有很大影响的问题ꎬ提出的 基于矩的步态周期检测方法中ꎬZernike 矩需要人体居中、尺度归一的前期预处理过程ꎬ而伪 Zernike 矩具有能描述运 动图像的特点ꎬ它可以避免人体居中、尺度归一等处理ꎬ以便直接测试步态的周期性.根据行走时的两帧之间的特征 取决于前一帧和后一帧的特征ꎬ提出了基于线性插值的矩阵步态识别算法框架ꎬ并且将投影特征、Hough 变换特征、 Trace 变换特征和 Fan ̄Beam 映射特征应用在 CASIA(B)步态库上ꎬ验证了框架的有效性ꎬ为解决步态识别问题带来 新的方法与思路.这种基于线性插值的矩阵步态识别特征本质上是一种权值不同的能量形式. 关键词:步态识别ꎻ矩阵步态识别ꎻ线性插值框架ꎻ步态周期检测ꎻZernike 矩ꎻ伪 Zernike 矩 中图分类号:TP391.41 文献标志码:A 文章编号:1673 ̄4785(2013)05 ̄0415 ̄11 中文引用格式:贲晛烨ꎬ张鹏ꎬ潘婷婷ꎬ等.线性插值框架下矩阵步态识别的性能分析[J]. 智能系统学报ꎬ 2013ꎬ 8(5): 415 ̄425. 英文引用格式:BEN Xianyeꎬ ZHANG Pengꎬ PAN Tingtingꎬ et al. Performance analysis of matrix gait recognition under linear in ̄ terpolation framework[J]. CAAI Transactions on Intelligent Systemsꎬ 2013ꎬ 8(5): 415 ̄425. Performance analysis of matrix gait recognition under linear interpolation framework BEN Xianye 1 ꎬ ZHANG Peng 1 ꎬ PAN Tingting 1 ꎬ WANG Kejun 2 (1. School of Information Science and Engineeringꎬ Shandong Universityꎬ Ji’ nan 250100ꎬ Chinaꎻ 2. College of Automationꎬ Harbin Engineering Universityꎬ Harbin 150001ꎬ China) Abstract:The existing gait period detection methods are not ideal and the performance of gait recognition is signifi ̄ cantly influenced by walking speed. Several novel gait period detection methods based on moments are proposed in this paper. The Zernike moment requires preprocessing including the assurance that the image of the human body is proportioned normally and is centered properlyꎻ the pseudo ̄Zernike moment may directly describe the motion im ̄ ageꎬ and it may avoid the need for such processing of making the image of the human body centered and sized nor ̄ mallyꎬ so as to directly detect gait periodicity. As the features of one frame are only decided by those of the prior and the rear frames in walkingꎬ a framework for a matrix gait recognition algorithm based on linear interpolation is proposed. Subsequentlyꎬ the projection featuresꎬ Hough transform featureꎬ Trace transform feature and Fan ̄Beam mapping feature are applied to the CASIA(B) gait database to prove the validity of the gait recognition framework. This brings new methods and understanding for solving gait recognition problems. This matrix gait recognition feature based on linear interpolation is essentially an energy form with different weighted values. Keywords:gait recognitionꎻ matrix gait recognitionꎻ linear interpolation frameworkꎻ gait period detectionꎻ Zernike momentꎻ pseudo ̄Zernike moment 收稿日期:2011 ̄10 ̄19. 网络出版日期:2013 ̄09 ̄29. 基金项目:国家自然科学基金资助项目(61201370)ꎻ高等学校博士学 科点专项科研基金资助项目(20120131120030)ꎻ中国博士 后科学基金面上项目(2013M530321)ꎻ山东省博士后创新 项目专项资金项目( 201303100)ꎻ山东大学自主创新基金 资助项目(2012GN043ꎬ2012DX007). 通信作者:贲晛烨. E ̄mail: benxianyeye@ 163.com. 步态识别是在人没有觉察的远距离情况下ꎬ通 过捕捉动态与静态信息来进行身份识别.由于步态 特征具有稳定性、普遍性、可采集性和惟一性ꎬ因此 可用于不同场合下的身份识别与认证[1] .Nixon 等[2] 总结了南安普敦大学 2002 年之前研发的步态识别 技术ꎬ同时预言了步态应用于生物特征识别的潜能. 2005 年ꎬBoulgouris 等[3] 首先将步态识别与其他生 物特征识别做了比较ꎬ提出步态识别可以作为多生 物特征识别技术的一部分ꎬ接着给出了步态识别系
向下翻页>>
©2008-现在 cucdc.com 高等教育资讯网 版权所有