Automatic calibration method based on improved camera calibration template

Camera calibration is one of the most important processes in computer vision. The systematic measurement accuracy largely depends on the accuracy of camera calibration. In this paper, we propose an automatic calibration method based on improved camera calibration template. For measuring object dimensions, we extract calibration points from the calibration template. However, there are many challenges in extracting additional information for camera calibration such as covering, nonuniform lighting, large viewing angle and distortion of camera. This new camera calibration template use new calibration points instead of the black and white squares on the traditional checkerboard. Because of this improvement, the automatic recognition algorithms can get a more accurate result than previous methods in identifying calibration points than previous methods. Experiments indicate that our novel approach outperforms most existing methods in some aspects such as scanning speed, stability and robustness. What´s more, it changes traditional camera calibration into an automatic process. At the same time, the prior information of the linear relationship contributes to determining the position of calibration points.