Improving the spatail resolution based on 4D light field data

This paper presents why the spatial resolution of an image captured by a plenoptic camera can be improved. The plenoptic camera captures the 4D light field (angular and spatial information of light) within a limited 2D sensor and results in reducing 2D spatial resolution due to inevitable 2D angular data. However, for improving the reduced resolution, we propose a novel analysis that 2D angular data contain spatially subpixel-shifted information. Although angular data have been defined as light's set radiated from one spatial point in a scene theoretically, practical angular data captured by plenoptic camera consist of spatially different minuscule area. That is, angular data provide the redundant data used generally by super-resolution techniques. Our experimental results demonstrate the improvement of spatial resolution as well as the existence of spatial information within 2D angular data.