Human Identification System Based on Gait using Active Horizontal Levels (AHLs) Feature and Chi-Square Attributes Selection (CSAS)

Nowadays, gait is a crucial field for many pattern recognition researchers. It is considered as a good way for biometric authentication in many surveillance systems. In this paper, a new method has been introduced to identify night walker images captured by infrared cameras. This method depends on the silhouette’s presence on different horizontal levels. A new proposed algorithm to select the Active Horizontal Levels (AHLs) feature has been presented. Chi-Square attributes (feature) selection and K-Nearest Neighbor classifier have been used to choose the AHLs that lead to the highest identification rate. The proposed method was evaluated against CASIA-C gait database, to recognize person from one image. The experimental results reveal the effectiveness of our proposed algorithm against other gait identification algorithms.