Vital Sign Integrated Tracking by Predictive KNN and Kalman Filter with UWB Radars

Ultra-wideband (UWB) radar is an effective tool for indoor tracking. By applying multiple radars in tracking, distance observations are obtained in different positions and the surveillance area is enlarged. However, even based on multiple radars, occlusions among targets still exist and cause loss of distance observations. Besides, imprecise initial positions introduce interferences at the beginning of tracking. This paper proposes a tracking approach to solve the problems. Firstly, the proposed approach integrates vital signs for target matching and provides precise initial positions for multiple targets tracking. Secondly, the proposed approach designs a predictive K-Nearest Neighbor method in both the raging stage and the data fusion stage, so as to reduce deviations in antenna-to-target distances and supplement the lost data. Experiments validate the effectiveness of the proposed approach in tracking multiple targets with a max average deviation 0.276m.