Depth Estimation Through a Generative Model of Light Field Synthesis
This work addresses depth estimation for computer vision tasks using light field photography, presenting an incremental improvement with a novel framework.
The authors tackled the problem of accurate depth recovery from light field data by proposing a generative model fully parametrized by depth maps, resulting in high-quality continuous depth map estimation through integration of regularization techniques like non-local means prior.
Light field photography captures rich structural information that may facilitate a number of traditional image processing and computer vision tasks. A crucial ingredient in such endeavors is accurate depth recovery. We present a novel framework that allows the recovery of a high quality continuous depth map from light field data. To this end we propose a generative model of a light field that is fully parametrized by its corresponding depth map. The model allows for the integration of powerful regularization techniques such as a non-local means prior, facilitating accurate depth map estimation.