Automatic real time gait recognition based on spatiotemporal templates

Gait recognition is a biometric method used to recognize humans based on the style of their walk. In the last few years, wide varieties of gait recognition approaches have been proposed, and significant improvements have been made. Unlike other biometric methods, such as face and body recognition, gait recognition requires dealing with a large number of video frames. As a result, most of the successful gait recognition algorithms are computationally expensive and not applicable for real-time surveillance applications. This paper focuses on developing a framework for automatic gait recognition, and proposes a novel algorithm to create a 2D spatiotemporal gait template that is reliable for person recognition in real-time surveillance applications. A neural network is used for classification where the input is the spatiotemporal gait template. The complete gait recognition framework developed in this paper involves automatic detection and segmentation of the human body, alignment and registration, feature extraction, and classification.