Biometric Gait Recognition with Carrying and Clothing Variants

Compact spatio temporal representation of human gait in form of gait enery image (GEI) has attracted lot of attention in recent years for biometric gait recognition. Researchers have reported very high recognition rates for normal walk sequences. However, the rates come down when the subjects are wearing a jacket or coat, or are carrying a bag. This paper shows that the performance for the variant situations can be improved upon considerably by constructing the GEI with sway alignment instead of upper body alignment, and selecting just the required number of rows from the bottom of the silhouette as inputs for an unsupervised feature selection approach. The improvement in recognition rates are established by comparing performances with existing results on a large gait database.