Investigation on the Recognition Validation Method of the Milli-Wave Radars for Automated Driving

Abstract

For automated driving and driver assistance systems, recognition performance in the real world is important to ensure safety. Cameras, millimeter-wave radar, LiDAR etc. are used for recognition, and their performance must be correctly evaluated. It is assumed that simulations will be used for the recognition performance of automated driving vehicles. However, the validity of simulations is greatly affected by the assumptions and premises of the model, so evaluations using actual objects are still considered to be important as a standard. Therefore, this study focuses on the evaluation of recognition performance of millimeter-wave radar in the 79 GHz band, which is expected to be implemented in automated driving, to estimate the recognition performance in the real world.