Calibration for Agricultural Vehicle Omni-Vision Localization System

Accurate calibration of a vision system is necessary for any computer vision task. A fast and practical calibration method of omni-vision system for agricultural vehicle localization system was developed. The calibration method utilized a 2D calibration pattern that can be freely moved. Without a priori knowledge of the motion, the boundary ellipse of the catadioptric image and field of view were used to obtain principal point and focal length. The extrinsic parameters were initialized via the homography from the calibration pattern to a virtual image plane. Calibration experiments and application experiments were done. Application experimental results show that the calibration improves the accuracy of the directional angles about 1° and improves the accuracy of localization about 1.6 cm in a 1.8 × 0.9-m square area. In conclusion, the calibration method is effective and important.