Spatio-Temporal Analysis For Gait Recognition: A Review

Abstract

Human gait recognition is an emerging research topic for analyzing human walking behavior which can be utilized in the machine vision fields. In video surveillance, the benefit of gait recognition is detecting humans in a distanced manner under the low-resolution videotapes. In recent years, detecting and tracking the abnormality of human gait is a challenging task. Spatial and temporal parameters are generally recognized as key metrics in Gait analysis. Spatiotemporal data describes the object states over time, an event or a position in space. The major goal of this review is to offer a wide-ranging overview of historical research on the gait prediction framework. This paper also describes the overview of existing gait databases, gait recognition representation used in the updated gait classification model. The extensive survey on Spatio-temporal parameters for improving the overall accuracy to predict the gait abnormality which is determined from the various gait databases.