Dynamic splitting method for clothing-invariant human gait recognition

Human gait as a behavioral biometric identifier has received much attention in recent years. But there are some challenges which hinder using this biometric in real applications. One of these challenges is clothing variations which complicates the recognition process. To mitigate this problem some approaches splits the silhouette into regions to isolate the negative impact of clothing in the affected region. In this paper a method is proposed for splitting the silhouette dynamically and according to the clothing type. The experimental results show that using this method increases the recognition rate significantly.