Gait recognition using artificial neural networks

Abstract

Gait recognition is the technique of identifying a person by the way they walk. person. For a small controlled environment the human identifying biometric methods such as fingerprint, face recognition, iris recognition requires co-operation from the subject. These biometric methods would not be sufficient for a large uncontrolled changing environment like surveillance where there is no co-operation of the subject. Henceforth special biometric namely gait is needed. Gait does not require the co-operation of the objects. Gait is new biometric in recent research. It identifies people at distance. This paper deals with an introduction to gait, advantages of gait over other biometric identification. It gives an outline to artificial neural network and discusses about Radial Basis Function Network. The methodology applied is: HMM is used to identify the parameters and RBF is used to training and testing process of gait recognition. In this methodology true positive rate is more than false positive rate which leads higher performance of gait recognition.