Group geometric calibration and rectification for circular multi-camera imaging system

Calibration on multi-camera imaging system in circular arrangement is a challenge task, because the preliminary shared feature points are absent among opposite viewpoints in the calibration procedure. In this paper, we propose a scheme of group geometric calibration and ideal rectification for circular multi-camera imaging system. In this scheme, a three-dimensional (3D) calibration object with new patterns is firstly designed to extract the extrinsic parameters. After that, the ideal camera projection matrix is estimated based on these parameters. Finally, we extend our scheme from static to dynamic scenario. In the experiments, our method has a superior performance over state-of-the-art calibration methods. Furthermore, we verify our scheme by application of 3D modeling and improvements can also be found. These results demonstrate that our method performance on such imaging system is applicable for future interactive visual applications.