LiDAR radar relative pose calibration

Abstract

A device for calibrating a LiDAR system (202) at a first position with a radar system (204) at a second position, comprising: A calibration target at a position and in an orientation with respect to the LiDAR system (202) and the radar system (204); a processor (206), configured to: Determining coefficients of a plane equation of the calibration target in a reference frame of the LiDAR system (202), Determining the coordinates of the plane of the calibration target in a reference frame of the radar system, Composing a cost function from a plane equation that includes the determined coefficients of the plane equation of the calibration target in the reference frame of the LiDAR system, the determined coordinates of the plane of the calibration target in the reference frame of the radar system, and a relative pose matrix that transforms the reference frame of the radar system to the reference frame of the LiDAR system, and Minimizing the cost function to estimate the relative pose matrix for calibrating the LiDAR system (202) with the radar system (204), wherein the calibration target includes a planar reflection area (212) and three or more angled reflectors (214), wherein the processor (206) determines the coefficients of the plane equation of the calibration target in the reference frame of the LiDAR system (202) using a LiDAR reflection signal received from the planar reflection area of ​​the calibration target, and determines the coordinates of the plane of the calibration target in the reference frame of the radar system from radar reflection signals received from the angle reflectors (214), wherein the calibration target is arranged at a plurality of positions and in a plurality of orientations and the processor (206) is further configured to: Determining the coefficients of the plane equation of the calibration target in the reference frame of the LiDAR system (202) for each of the plurality of positions and orientations; Determining the coordinates of the plane of the calibration target in the reference frame of the radar system (204) for each of the plurality of positions and orientations; Assembling a summed cost function to include a summation of cost functions for each of the multitude of positions and orientations, each cost function containing the determined coefficients, the determined coordinates, and the relative pose matrix for the associated position and orientation of the calibration target; and Minimizing the summed cost function to estimate the parameters of the relative Pose matrix, wherein the LiDAR system (202) and the radar system (204) are arranged on a vehicle (100), wherein the processor (206) is further configured to track an object (104) in relation to the vehicle (100) using the LiDAR system (202), the radar system (204) and the estimated relative pose matrix, wherein the object (104) is a lamppost.

Images