A method and system for calibrating extrinsic parameters of a camera and a lidar sensor device

Abstract

The application discloses a camera and laser radar sensor device external parameter calibration method and system, first, using a collection device to collect images and point cloud data, road surface extraction is carried out in the original point cloud data collected by the laser radar, the extracted road surface point cloud data is obtained by using the voxel completion mode to obtain the road surface point cloud completion data, and the reflection intensity value of the completed point cloud data is histogram equalized to enhance the contrast of the lane mark and the road surface reflection intensity; then the road surface point cloud data is projected to the image plane and matched with the camera image, and the obtained matching feature points are used to solve the initial external parameters; finally, the mutual information value is used to construct a target function composed of elements in the external parameters, and the optimization algorithm is used to optimize the best external parameters. The method has the advantages of ensuring the calibration accuracy, being capable of completing the calibration on line, being simple in process, being strong in adaptability, and meeting the online calibration demand of the external parameters in the automatic driving system.

Images