Augmented Reality on a Deformable Surface by Considering Self-Occlusion and Realistic Illu

Abstract

In this study, we augment a virtual object on an image of a deformable surface, such as a piece of paper, which is acquired from a web camera. To increase presence, we consider self-occlusion and realistic illumination of the augmentations. To determine the geometric relation between the camera and the deformable surface, we use markers that are printed on the surface. Using prior marker information, we can obtain the homography and calculate three-dimensional coordinates of the surface. If some markers cannot be detected because of self-occlusion, we estimate the occluded points by using a Kalman filter. We try to increase presence by applying realistic illumination to augmented textures.