Camera External Self-Calibration for Intelligent Vehicles

摘要

An algorithm for the camera external self-calibration is presented in this paper which is quite convenient and practicable as no calibration board is required. The camera is fixed rigidly on a vehicle with a planar motion constraint. Based on that, the rotation parameters are decomposed into two parts and the vehicle is manipulated to take two certain motions to solve those two parts in turn. Furthermore, given that the velocity is available, the translation can be solved by introducing a virtual aligned camera, which can improve the accuracy and reduce the computational time. Both simulations and experiments are carried out where several relevant algorithms are adopted as comparisons. Corresponding results validate the performance and handiness of the proposed method.