Human gait recognition based on motion analysis including ankle to foot angle measurement

摘要

This paper proposes an intelligent authentication of a human using various feature extraction. In this methodology, the silhouettes are pre processed to remove the noises to attain the extraction of various angles, including ankle to foot. From the head to foot almost six angle values are calculated using skeletonization method. Height and width of the silhouettes are also measured. Using the multi class SVM the effective human gait recognition is obtained.