Gait Recognition Based on Static and Dynamic Features Fusion of Body Biometrics

A simple and efficient motion-based gait recognition algorithm was proposed by spatial-temporal silhouette analysis.For each image sequence,a background subtraction algorithm was first used to segment and track the moving silhouettes of a walking figure from the background.It was attempted to describe gait contour by using the distance of the lower limbs on the edge,then to make quasi-periodic analysis of the foot width on the gait image and to solve the problems resulting from image sequence of different gait cycle by using fusion of hidden Markov model and static features of body biometrics at a decision level.By utilizing the algorithm,the experiments with Little Boyd,NLPR and CMU databases have achieved comparatively high correction identification ratio.