INDIVIDUAL GAIT RECOGNITION USING PARTICLE SWARM TEMPLATE SEGMENTATION

Abstract

Human identification by GAIT recognition now a day’s dealing with great challenges in biometric
technology. In this paper, the Particle Swarm Template Segmentation method is proposed to identify
the person using their GAIT. The PSTS method employs the Particle Swarm Optimization algorithm,
which is a relatively recent metaheuristic population-based swarm intelligence optimization technique.
A comparative analysis has been done on existing and proposed methods to evaluate the performance
of Particle Swarm Template Segmentation (PSTS)over another metaheuristic evolutionary based
genetic Template segmentation (GTS),using Gait Energy Image and Gradient Gait Energy Image
templates of gait sequences from CASIA – B (Chinese Academy of Sciences) dataset. Further Principal
Component Analysis and Linear Discriminate Analysis are used for accuracy by reducing
dimensionality in our data in GAIT recognition. The results depict that the for individual gait
recognition, the Gradient Gait Energy Image (GGEI) appears to display the best outcome when
segmented with the Particle Swarm Optimization algorithm compare to Genetic Template Segmentation
(GTS)