Robust Intrinsic and Extrinsic Calibration of Multiple Camera with a One Dimension Object

Camera calibration is essential in visual measurement. This paper proposes a robust 1D calibration method for multi-camera systems. Compared to existing algorithms, this paper introduces a new fundamental matrix algorithm that reduces the condition number of matrix computation while removing potential outliers, thereby improving the accuracy of the fundamental matrix calculation. Following this, a hierarchical reconstruction method is proposed, which allows for the joint calibration of multi-camera intrinsic parameters and their relative poses with only the free movement of a 1D calibration object. Both simulation and real experiment results consistently demonstrate that this method not only enhances the accuracy and robustness of multi-camera intrinsic and extrinsic parameter calibration but also ensures the simplicity and efficiency of the calibration process. This makes it suitable for high-precision multi-camera joint calibration in various scenarios.