Wearable Gait Recognition Systems Based on MEMS Pressure and Inertial Sensors: A Review

Abstract

Gait is a basic characteristic of human motion. Different gaits are usually associated with different body functions. Gait recognition has wide applications in clinical medicine, rehabilitation training, posture recognition, and other fields. At present, many wearable systems based on MEMS pressure and inertial sensors have been developed and used for gait recognition. However, there has been a lack of a comprehensive summary and comparison of these systems from the perspectives of their hardware compositions, working principles, algorithm models, and applications. This review aims to promote the development of wearable gait recognition devices. First, sensor technologies for gait measurement are summarized from the perspectives of their working principles and characteristics. Then, these technologies are compared in terms of the performance of algorithms for data preprocessing, cycle and phase segmentation, and gait recognition. Next, the applications of MEMS sensor-based gait recognition systems in various fields are summarized. Finally, some limitations of existing wearable gait recognition systems and their future directions are discussed.