Model-based gait tracking method: A review of recent development gesture interaction

Gesture interaction by gait is becoming a topic of interest in human computer interaction due to its advantage of identifying individuals. This paper provides a brief review of recent developments on gait tracking and gives an overview of model-based approaches for gesture recognition. To recognize gait gesture in real-time, model-based approach is generally used in gesture interaction. The camera viewpoint provides reliable geometric information for analyzing gait gesture. Thus, gait posture can be estimated by referring to projective relationship between a model and a camera. This survey has three major issues involved in gait image representation, feature tracking, and pose correction. The conclusion of this paper includes some discussions of future work and provides addressing directions for the domain.