多特征和多视角信息融合的步态识别  

Gait Recognition Via Multiple Features and Views Information Fusion

作　　者：赵永伟[1] 张二虎[1] 鲁继文[1] 胡峻林[1] 

机构地区：[1]西安理工大学信息科学系,西安710048

出　　处：《中国图象图形学报》2009年第3期388-393,共6页Journal of Image and Graphics

基　　金：陕西省自然科学基础研究计划项目(2006F26);陕西省教育厅自然科学研究项目(07JK353)

摘　　要：提出了一种基于多特征和多视角信息融合的步态识别方法。应用背景差分和阴影消除获得人体步态轮廓,对人体轮廓使用伪Zernike矩、小波描述子和Procrustes形状分析法进行了特征提取。通过多特征和多视角步态信息融合,完成了基于人体步态特征的身份识别。该方法在CASIA步态数据库上进行了实验,取得了较高的正确识别率,实验结果表明本文所提出的识别方法具有较高的识别性能。A new gait recognition method based on information fusion of multiple kinds of features and views is proposed in this paper. Through the background subtraction and shadow elimination, human motion silhouettes are obtained and gait features are extracted using pseudo-Zernike moment, wavelet descriptor and Procrustes shape analysis. The gait recognition is accomplished through information fusion of muhiple kinds of features and views on feature level and decision level. The method is evaluated on the CASIA gait database and received comparative high correct recognition rate. The experimental results show that our approach has efficient recognition performance.

关 键 词：步态识别 伪Zernike 矩 小波描述子 Procrustes形状分析 信息融合 

分 类 号：TP391.41[自动化与计算机技术—计算机应用技术]