Utilizing gait traits to improve e-border watchlist performance

The objective of this study is to develop a framework to quantify gait characteristics that can be used in biometric-enabled watchlist technologies. One of such applications is a border control passage equipped with sensors that are used for acquiring biometric traits. As a proof-of-concept, we consider gait trait data which is collected in real-time and used to classify certain gait patterns. Specifically, we aim at identifying gait abnormalities such as a limp in either leg, which often exists as metadata in watchlists or forensic records. This paper examines a few techniques for gait analysis, as well as feature extraction and classification aimed at distinguishing normal gait from abnormal gait. The concept of expanded watchlist technology is explained, in which the gait characteristics can be used in an inference engine for risk assessment in biometric-enabled border control.