| It is one of the main tasks of vision navigation system in autonomous vehicle to give the position information of the road and obstacles in the enviroment. The obtaining of the position information needs camera-on-vehicle's parameters. This dissertation researches on An Automatic Extrinsic Parameter Calibration Method for Vision Navigation Camera which is equipped on Autonomous Land Vehicle (ALV) on highway. A novel automatic camera extrinsic parameter calibration algorithm is proposed and proved in the HONGQI prototype.The describing form of road lines on highway is researched first. Then the parameterized description of horizontal road on highway is proposed.To illustrate the issue of calibration and automatic calibration of the camera, this dissertation researches on the particularity of camera calibration in vision navigation process of ALV as well as the possibility of automatic calibration in the process.According to the parameters relationships between road models in both the image and vehicle coordinate, this dissertation reseaches on the combination of vehicle kinematic equation,road models in two coordinates and changing of camera's pose through Extended Kalman Filter(EKF).Parameter values are predicted by EKF. So that automatic camera calibration of the main part extrinsic paramets is accomplished.Finally, this method is verified by Simulation and in HONGQI ALV. Experiments results show that this method has good performance in the structural scenery on highway. |
PDF Full Text Request