Cumulative Clustering Filter for MIMO Radar Detecting Human Hand Intrusion

MIMO radar sensors with a 3-dimension detection zone are recently used for safety-related sensors. They may possess the capability of detecting human body parts such as hands. When the reflectivity of a target is low and the sensitivity of the radar is therefore relatively high, it is necessary to eliminate cluttered noise out of detected points and increase the possibility of detecting the target at the same time. This paper presents a data processing method including a cumulative filter and clustering, for achieving the goals above. Taking the human hand intrusion scenario as an example, an experiment was conducted to investigate the improvement of the detection performance by implementing this cumulative clustering filter.