Abstract: An important task for object tracking and positioning is camera calibration, whose purpose is to estimate the internal and external parameters of cameras. A exact mathematical model, which could provide a fast method to obtain the camera centre, was established based on a strict geometric constraint relationship. The other parameters of the camera was acquired in a resonable resolving sequence to guarantee the parameter accuracy of the camera. The pixel position in a distorted image is corrected based on the computed calibration parameters to recover the original space relationship between pixels, then the accurate and undistorted images are obtain. After the position of the spot on the calibration template is computed based on camera calibration parameters, the computed spot position is compared with its real position for verifing its accuracy. The experiments to check the image correction effect and accuracy indicates that the calibration parameters obtained by the algorithm are accrate and reliable, the accuracy is effectively improved, and the accuracy requirement for the position estimation of the target in moving-target tracking can be satisfied.