Carcorder Camera Calibration method of external parameters based on lane line

Abstract

The calibration of the camera’s internal and external parameters is the basis for constructing a three-dimensional world from a two-dimensional plane. It is also the basis for corresponding point matching and image stitching. The accuracy of camera parameter calibration directly affects the accuracy of the restoration of the three-dimensional world. The internal parameters of the camera have been fixed since the factory, no need for frequent calibration, but the external parameters of the camera vary with the installation position of the camera. For the driving recorder, it needs to be re-calibrated after each repositioning. Therefore, it is particularly important to find a simple, efficient and high-precision external parameter calibration method. The calibration method based on chessboard grids or control points has the disadvantages of large amount of engineering, complicated operation, low efficiency, etc. Based on the geometric characteristics of roads with fixed width and parallel lane lines, this paper proposes a solution for external camera calibration using 3 sets of parallel lane lines within flat roads. The method of calculating the external parameters of the on-board camera is to establish conditional constraint equations by selecting equally spaced control points on different lane lines, and then use the least square method to calculate the optimal value of the rotation matrix from the camera coordinate system to the vehicle coordinate system, the height of the camera to the ground, and of the x and z coordinates of the lane line control point in the vehicle coordinate system simultaneously. The method is simple and easy to implement, and the calibration accuracy meets the requirements of high-precision lane-level navigation map production, combining with the vehicle coordinates measured by GPS and the lane line coordinates under the vehicle coordinate system, the lane lines under the earth coordinate system can be solved, providing a basis for the calibration of external parameters of the driving recorder and the production of high-precision maps.