Chaotic characterisation of frontal normal gait for human identification

Human recognition using gait features in predominantly frontal-normal motion has been described in this paper. Compared to current methods for gait identification, this allows convenient combination of other biometrics using a single camera. We analyse how this motion yields more dynamic information, allowing us to characterise gait in a new way, using nonlinear dynamics of time series normally used in chaos theory. Using chaotic measures to identify humans by their gait is a significant precedent. Phase-space analysis of trajectories of a set of Moving Light Displays (MLDs) provides sufficient information for identification of people by their gait. A number of experiments has been set up to demonstrate the viability of this approach which contribute to the relatively unexplored area of fusion of face with gait. This provides a more robust identification scheme.