Fish-eye cameras calibration for vehicle around view monitoring system

This paper presents a camera calibration method for estimating the intrinsic parameters of fish-eye cameras which would be utilized in the vehicle around view monitoring system. A calibration pattern is devised to provide the information for calibrating all of the fish-eye cameras simultaneously. The diagonal projection is applied to extract the feature points on the calibration pattern. The lens distortion coefficient of a fish-eye camera can then be obtained based on the boundary of extracted feature blob. The camera lens principal point is evaluated by using the symmetry of the dark zones around image corners. The extracted feature points on the calibration pattern could further be utilized to generate the bird's-eye view image around vehicle.