The light field images capture both spatial and angular information of a scene. This enables the user to freely navigate a scene to some extent. Another benefits include refocusing in frequency domain. Light field reconstruction problem aims to use the redundant information between Light Field Images to reconstruct a dense light field given a sparse input images. This enables to represent a scene with just 2x2 or 3x3 images instead of 7x7 or 9x9 images.
Light field reconstruction can minimize the data requirement for representing a scene thus enabling use of light field in applications which require real time computing, for example VR.