文章基本信息

标题：Background Subtraction and Target Classification for Gait Recognition
本地全文：下载
作者：Kantipudi MVV Prasad ; Dr.V.Sailaja ; A. Jagan 等
期刊名称：International Journal of Computer Technology and Applications
电子版ISSN：2229-6093
出版年度：2011
卷号：2
期号：3
页码：515-520
出版社：Technopark Publications
摘要：This paper deals with background modeling and moving object classification for gait recognition. Current image-based human recognition methods such as fingerprints, face, iris biometric modalities, generally require a cooperative subject views. These methods cannot reliably recognize non cooperating individuals at a distance in the real world under changing environmental conditions. In such conditions, recognition of a person using gait has good advantage. First step in gait recognition is the background subtraction/modeling. This is the crucial step in gait recognition. By using this, identification of moving objects from background scene has to be done. Perfect background subtraction is essential to get a high recognition rate. Next step is the separation of human beings from other moving objects (viz; car, tree etc.). In this paper, we have used a modified background subtraction algorithm and subsequently used feature-based classification of pedestrian from other moving objects. Experimental results demonstrated the effectiveness of the proposed method
关键词：Gaussian Mixture Model; Temporal consistency; Classification metric operator.