A calibration method of camera with large field-of-view based on spliced small targets

Camera calibration is an important process in computer vision. For the issue that a large scale target is hard to fabricate, while using a small target can only get the local optimal results in the calibration of camera with large field-of-view, a calibration method based on spliced small targets is proposed. In this method, small targets are spliced into a large scale target through camera polynomial projective model using difference solution. High precision calibration of camera with large field-of-view is then conducted with the spliced large scale target through traditional calibration approach. At the same time, in order to eliminate the restriction that small targets only can translate, not rotate when using difference solution to splice them, a method to solve the angle between two straight lines based on invariance of cross ratio is proposed, while respective composite target is designed, which makes the method proposed in this paper more flexible. Feasibility and stability of the method in this paper are verified through simulation and experiments. The experiment results show that the method proposed in this paper can achieve the high precision calibration for camera with large field-of-view.