Use of ArUco markers for image registration of photographs and influence of camera tilt on process performance

Abstract

Image registration is widely used in the field of graphic technology, from printing processes to research. In order to successfully align two images, image sections must be recognized. In printing, simple linear markers can be used for alignment, while advanced visual systems use fiducial markers, such as ArUco markers, for specific position recog- nition. These can be applied to any surface and are quickly recognized by open-source algorithms. This gives us the ability to recognize any visually perceived surface and align it with the target image. This research focuses on the applicability of ArUco markers as reference points for image registration of photographs of flat, uniform surfaces. The camera tilt position was adjusted in four combinations to mimic the imperfect positioning of the camera relative to the observed surface. Twelve marker combinations were tested, varying in size, number, and positioning. Research has shown that ArUco markers are suitable as reference points for image registration, but in specific combination of size and position. General guidelines for their use have been established.