Photometric compensation for practical and complex textures

Abstract

We propose a photometric compensation for projecting arbitrary images on practical surfaces of our everyday life. Although many previous proposals have achieved fine compensation at their experimental environments [Nayar et al. 2003], they cannot support practical targets including high-contrast texture. In order to adapt to such situation, we need a time-consuming iterative processing with camera feedback. Even though the iterative processing is applied, we cannot obtain fine compensation because no camera pixels of a projector-camera system (procam) correspond perfectly to the pixels of the projector [Mihara et al. 2014].