Gait Classification of Healthy Young and Elderly Adults Using Micro-Doppler Radar Remote Sensing.

This paper presents a radar-based technique for the gait classification of healthy young and elderly adults based on the difference in their balance ability. We propose a method for extracting the velocity parameters related to the balance ability during walking via micro-Doppler radar gait measurement. The experimental results show that significant gait differences between young and elderly adults exist in the extracted parameters. In particular, the effectiveness of the parameters related to leg motions in both stance and swing phases is verified. By using the extracted parameters, a support vector machine-based approach shows that two groups can be effectively classified.