Gait Based Authentication with Hog Feature Extraction

Human Gait, a soft biometric helps to recognize people by the manner, they walk, it also used for recognizingphysical behavioural of human. The process of human motion analysis and biometrics has attracted several secured environments such as military, banks and airports etc, which leads to be a popular research direction. In this paper the human behaviour is recognized from a set of video samples and the features are extracted using HOG transform. KNN classifiers are used to classify the features extracted from the videos. The HOG feature based analysis has achieved higher recognition and accuracy of 93%compared to the existing methods.