Gait Biometric Recognition Using the Footstep Ground Reaction Force

Abstract

The design of a gait-based biometric system is highly dependent on the aspect of gait being studied and may be limited by the tools available for study. The research presented in the previous chapter has suggested that the ground reaction force , measured using either a WS or FS approach, may lead to better results than might be achieved using a more conventional MV approach. Consequently, for the purpose of the analysis provided in this book, we have chosen to study GRF-based gait recognition, and for the remainder of the book, we will be performing the analysis on the GRF data acquired via a floor-mounted force plate (an FS approach ). The GRF is typically represented as one or more discrete-time signals measuring the force exerted by the ground back on the foot at varying points in the footstep. This signal representation of the GRF is advantageous because it opens the gait biometric for study using well-established and leading edge analysis techniques, which have often previously catered to the examination of similar problems in alternate domains. In this chapter, we begin by exploring the GRF in greater detail and proceed to describe the varying analysis techniques used in previous GRF recognition studies. In doing so, we identify several key research gaps, which will be addressed via novel research presented in the chapters that follow.