Gait Recognition Using Spatio-temporal Templates and Local Moments
Vision-based human identification at a distance in surveillance has recently gained more attentions. Gait has the advantages of being non-invasive and difficult to conceal, and is also the only perceivable biometric at a distance. In this paper, we propose a novel framework for gait analysis and recognition applications. In order to create a simple and compact gait sequence representations, we construct spatio-temporal templates, called Stride History Images (SHI). In order to delme the stride length and ensure an invariant analysis due to differences in gait period,we perform a gait period estimation procedure using nonlinear analysis beforehand. In order to present a desirable descriptor of feature distributions, we partition the SHI images and find a suite of feature vector sets created from local moment invariants.In a further processing stage these feature vectors corresponding to a gait sequence is used for identification. Despite its low computational cost, the proposed approach achieves recognition capability by an 85.57％ CCR on Soton database and concrete improvements are seen in comparison to other methods of similar or higher complexity.
gait recognition spatio-temporal templates stride history images, moment invariants, local moments
CHEN Shi GUO Qiuli GAO Youxing
Zhejiang Wanli University Ningbo, Zhejiang Province 315100, China Computer Peripheral Institute XIDIAN University Xian,Shaanxi Province 710071,China
国际会议
武汉
英文
631-635
2007-07-25（万方平台首次上网日期，不代表论文的发表时间）