Calibration of Position and Orientation between Cameras without Common Field of View Using Cooperative Target

In online surface settlement monitoring system of image-based ballastless track, a transfer station consisting of a multi-camera without a common field of view is often used. Because measurement error will accumulate as the number of transfer increases, position and orientation relation betweeen cameras should be calibrated. Traditional calibration methods based on total stations are pretty complicated. We found that the hand-eye calibration of the robot is equivalent to this problem. Firstly, the multi-camera is moved twice in small step and the multi-camera captures the cooperative target image at three different positions and image physical coordinates of four mark points in each image are extracted. Secondly, in order to solve the extrinsic parameters of the camera, we use a special P4P algorithm for feature points distributed in a square. Lastly, the matrix rearrangement method is used to solve the hand-eye transformation matrix according to the extrinsic parameters of the cameras. The experimental results show that the measurement accuracy of this method is same as the traditional method. Moreover, the method features simple operation, less calculation task and high precision position and orientation.