关键词:步态识别; 步态能量图; 核主成分分析; 支持向量机
GEI based gait recognition by using KPCA and SVM
LIANG Shao-cong, ZHOU Ming, LI An-an
(College of Computer Science, Sichuan University, Chengdu 610065, China)
Abstract:This paper present a novel approach to extract gait features based on gait energy image (GEI). GEI was obtained by analysing the gait circle form the size normalized silhouette image sequence. Since PCA’s limite ability to deal with the nonlinear data, applied the kernel principal components analysis (KPCA) method to reduce data dimension, and then performed gait classification and recognition by support vector machine by its good performance in the classification problem. The method is evaluated on CASIA gait database and the result of the experiment demonstrates that the approach has greatly improved the recognition performance.......