Extrinsic parameter calibration of 2D radar-camera using point matching and generative optimization

Abstract

The fusion of sensors is a task to improve the recognition performance of obstacles in an autonomous vehicle system. For the fusion of sensors, extrinsic parameter calibration with 6-dof is needed to integrate the data between the sensors. This paper proposes a method of calibration between 2D radar and camera using point matching. We use corner reflectors calibration target, which focuses the radar signal at the center of the target. Using this target, we set the center point as Ground truth, which improves the accuracy of the estimation of the calibration parameter. We simultaneously measure the target with radar and camera. After estimating the intrinsic parameters of the camera, the coordinate axes of the camera and the radar are matched based on the camera view. This paper estimates the extrinsic parameters using point matching and top-down method. The proposed method is verified through experiments and results.