A Study on the Motion Analysis of Indoor Moving Object Using mmWave-FMCW Radar

This paper presents a study on the classification of moving objects in indoor environments using mmWave-FMCW radar and point cloud data. The proposed method combines k-means clustering and autocorrelation peak strength analysis to distinguish between periodic and non-periodic motions, specifically targeting fans and curtains as common sources of false detections in smart homes. The experimental setup includes scenarios involving a single fan, a single curtain, and multiple objects (two fans and a curtain) within the same space. Results demonstrate the ability of the system to segment point cloud data into distinct clusters and accurately classify objects based on their motion characteristics. The periodic motions of fans were identified with high peak strengths, while the non-periodic motion of the curtain exhibited lower peak strength. The findings validate the effectiveness of the proposed approach in reducing false detections and enhancing object classification, paving the way for robust motion detection systems in smart home environments.