Human gait recognition based on signals from two force plates

Abstract

This paper presents a practical approach to person recognition by gait. The biometric system is based on ground reaction forces and positions of the center of pressure generated during the subject gait and the single-output multilayer neural network. The article discusses both the identification and the verification problems as well as influence of the security level on the quality of the proposed biometric system. The achieved results (almost 96% of correct recognition) show that human gait is a biometric measure which enables efficient authorization in a simple way. © 2012 Springer-Verlag Berlin Heidelberg.