Study on condition constraints of classical camera calibration

Camera Calibration is a crucial task in computer vision. It aims to get internal and external parameters of camera from images. Currently, classical calibration techniques are widely used in measurement and monitor, especially the method proposed by Zhang Z.Y. In this paper, it is demonstrated that classical method is rather inaccurate in some certain conditions-different distance. Experiment results on real images are analyzed. Meanwhile the method is carried out for hundreds times with different numbers of images to point out the inaccuracy of the method involved with iterative algorithm. This work provides the basis for further development in camera calibration, and plays a significant role for close-range calibration in engineering.