Universal calibration for a ring camera array based on a rotational target

Abstract

Panoramic dynamic and static measurements of objects in the application of vision measurement are difficult due to the constraints of a camera field of view and multicamera calibration technology. This paper proposes a universal global calibration method for ring multicamera systems based on rotating target and multi-view vision technology. This method uses a rotating target to establish the relationship between ring multi-camera arrays, retrieves the coordinates of the target corners from the fields of view of different cameras, and combines them with the rotation angle to complete the coordinate unification of the system. The unification of coordinates is unaffected by the overlapping fields of view between cameras, and the number of cameras can be configured arbitrarily. The calibration accuracy, validity, and precision of the proposed method are verified through reprojection error, dynamic tensile test, and 3D reconstruction.