A Computer Vision Based Testbed for 77 GHz mmWave Radar Sensors

In this paper, our main objective is to develop a tool (testbed) that allows real-time testing and visualization of mmWave radar algorithms using a high-level programming language. This tool is capable of capturing data from the radar module without any loss of packets and efficiently display results as heatmap images of target distance and velocity. This new testbed is especially tailored for hardware-in-the-loop (HIL) type of experiments involved in Advanced Drivers Assistance Systems (ADAS) and Autonomous Vehicles (AV) among others. To validate the efficacy of the developed testbed, Texas Instruments (TI) AWR1642 module was used. Results show that the test bed described hereafter is successful and the proposed hardware and software configuration will allow more advanced image processing and computer vision techniques to be used for a variety of applications. Several examples are shown with Python being the main programming language for control and data acquisition and openCV for visualization of radar images.