الفهرس 
PROPOSED TECHNIQUE FOR ENHANCING GAIT RECOGNITION ACCURACY
Post processing and TrackingOnly 14 pages are availabe for public view
 |  | from | 109 |  |  |
| | | from | 109 | | |
Abstract
Non-interactive biometric systems have gained an enormous interest among computer vision researchers as they provide more efficient and reliable ways of identification and authorization from a distance. Face and gait recognition are types of non-interactive biometric systems without users cooperation with the surveillance system. Gait recognition is more productive than face recognition as it can manage low resolution and low brightness images. It aims to know the individuals based on their style and the way of their walk. Yet, gait recognition performance is frequently deteriorated by some variety of factors, such as a viewing angle variation, clothing and carrying condition changes. Recently, deep learning models can be employed efficiently in the gait recognition system. Such models are a class of machines that can take on various layers of feature hierarchies and automatically assemble high-level features from low-level ones. Hence, they are more generic since the feature construction process is completely robotized. A standout amongst the most utilized deep learning models is the Convolutional Neural Network (CNN) architecture.This thesis proposed a technique to establish an automatic gait recognition system namely, “OFGEINet” based on spatiotemporal silhouette analysis measured during walking. It comprised three vital modules, specifically, 1) human detection and tracking, 2) feature extraction and 3) training or classifying. Accordingly, the first module serves to identify the walking figure in an image sequence.