Research on gait-based human identification

Abstract

Gait recognition refers to automatic identification of individual based on his/her style of walking. This paper proposes a gait recognition method based on Continuous Hidden Markov Model with Mixture of Gaussians (G-CHMM). First, a Gaussian mix model is initialized for training image sequence with K-means algorithm, and then training the HMM parameters using Baum-Welch algorithm. These gait feature sequences can be trained and obtains a Continuous HMM for every person; therefore, every person's gait sequence can be represented by the 7 key frames and HMM. The experiments, utilizing CASIA gait databases, present a comparatively correction identification ratio and a comparatively robustness when the bodily angle varying.