ABSTRACT
With the increased interest in security and surveillance, the effective authentication of people has assumed great significance in recent years. Consequently, the development of highly accurate biometric authentication systems has received prominent attention in the research community. Biometric authentication is defined as the process of utilizing unique personal or biometric features to identify a person. Popular biometric features used for person authentication include the face, fingerprint, DNA, retina, etc. However, the use of these features in biometric authentication systems requires the active cooperation of the subject involved, which is often difficult and inconvenient. To avoid the need for active user cooperation, without compromising on the authentication accuracy of the biometric authentication system, researchers have investigated the use of human gait, or walk, as a biometric feature. Specifically, the property of human gait or walk being unique to each individual has been utilized in biometric authentication systems to identify or verify individuals. In this chapter, we present a detailed overview of a gait-based biometric authentication system and present a discussion on the various components of the system. Additionally, we categorize the literature on gait-based biometric authentication systems into three classes: camera-based systems, floor sensor-based systems, and wearable sensor-based systems, and present a detailed comparative study. Finally, we present a short discussion on the susceptibility of gait to mimicking attacks, thus motivating the need for gait-based multimodal biometric authentication systems.
