Development of Kalman Filter-Based Tracker and Visualizer for Millimeter-Wave Radars in Python

In this paper, we present a Kalman filter (KF)based tracker and visualizer in a tightly-coupled architecture for millimeter-wave (mmWave) radar applications. For the tracker segment, we consider an interactive multiple model (IMM)based KF framework, which combines constant acceleration and velocity models. The visualizer unit uses a multi-thread approach ensuring thread safety, and can be implemented on software with limited memory and computational power. Moreover, the tracker and visualizer unit can be run on different computers due to the TCP/IP socket connection. The tracker-visualizer duo is developed in Python, as a scalable solution, which makes it platform-agnostic. Through real-world experimental data, we demonstrate the efficacy of the proposed tracker and visualizer.