摘要： 步态识别通过人体走路的姿势来识别人的身份。近年来，步态作为一种生物特征识别技术备受计算机视觉研究者的关注。对某个人的一个步态序列利用动态Viterbi算法得到一个样本姿态序列，对其多个步态样本姿态序列的对应姿态取平均得到这个人的特征姿态序列，对特征姿态采用主成分分析法和线性判别分析法处理特征空间，并用最近邻法进行识别。利用CASIA数据库对本文方法进行验证，取得了较高的识别率，并对体形变化具有较强的鲁棒性。

关键词: 步态识别, 隐马尔可夫模型, Viterbi算法, 主成分分析, 线性判别分析

Abstract: Human gait recognition is the process of identifying individuals by their walking manners. The gait as one of biometrics has recently drawn attention to the computer vision researchers. The stance estimation of one’s each gait sequence is based on the dynamic programming-based Viterbi algorithm, which returns a sample stance sequence. The corresponding stances of sample stance sequences are averaged to get his feature stance sequence. Principal component analysis and linear discriminant analysis are used for feature transformation of feature stances in the feature stance sequence. The recognition is achieved by nearest neighbor algorithm. The experiments made on CASIA gait databases obtain comparatively high correction identification ratio and comparatively strong robustness for variety of bodily form.

Key words: gait recognition, Hidden Markov Model(HMM), Viterbi algorithm, principal component analysis, linear discriminant analysis

中图分类号: 

• TP391