Investigating radar-based micro-motion signatures for human detection and identification in short range indoor environments

Abstract

The study of radar for human recognition based on deep learning is becoming increasingly popular. It has been demonstrated that the micro-Doppler (μ-D) spectrograms effect can reflect walking human gait by capturing the periodic micro-motions of the limbs. The research scope was extended to include human recognition for variable activities, and hence, a broad number of applications have been investigated, such as fall detection. In addition to this, there are two main factors that have introduced the radar as a powerful sensor for such applications. First is the radar detection capability that is not affected by any environmental limitations. Second, the multiple-input-multiple-output transmission protocol that enabled the radar detection and tracking for multiple humans.