CCD camera calibration technique is an important part of image processing. The calibration accuracy of CCD camera parameters determines the precision of the image processing system. With the development of industry, the demand for image processing precision is increasing, so the precision and robustness of CCD camera calibration technique are required to be higher. In order to further improving camera calibration accuracy and simplifying calibration process, in this paper, we proposed a new method of camera calibration on the basis of the traditional camera calibration technique. Different from the traditional calibration method, this new calibration technique uses a "black box system" to build a link between the ideal image points and actual ones instead of setting up linear CCD camera imaging model (i.e. pinhole imaging model) and nonlinear imaging model about camera distortion. The actual pixel coordinates of the camera image are the input values, and the reprojection points obtained by the world coordinates are the output values. In order to verify the validity of this method, we carry out experiments based on the improved calibration method of zhang zhengyou.1 And finally we verified the feasibility of this method by comparing the experimental results respectively obtained though carrying out the calibration method this paper proposed and the traditional calibration method by camera calibration toolbox. © 2018 SPIE.