Acquisition and processing of Ray-Space/Light Field data

Ray-Space/Light Field is a novel representation of 3-D visual information that enables 3D/free-viewpoint photo-realistic image generation. Although it is based on a clear and simple concept, and hence, it is expected to be a basic representation for 3D end-to-end systems, it has the problem that there exists no easy way to obtain Ray-Space data efficiently. In this paper, we first describe the basic concept of the Ray-Space and then review several acquisition methods, including a high number of multi-camera systems and space/time division acquisition methods. Among them, we focus on a Light Field Camera (LFC), which is capable of obtaining 4-D light ray data in a single shot using the space division method. We point out that one of the drawbacks of this system is the low resolution of the reconstructed images and introduce our approach to solve this by using a super-resolution method. Some preliminary experimental results are provided in the paper.