Underwater Image Dehazing with a Light Field Camera
K. Skinner · M. Johnson-Roberson
TL;DR
A light field camera lets underwater images be dehazed by combining per-view 2D dehazing with guided filtering, tested on a public water-tank dataset.
BRIEF
Light scattered by particles in water produces a hazy veil over underwater images, much like fog in air. Fixing this from one image is ill-posed and usually relies on strong assumptions about the scene, while multi-image approaches need extra cameras, polarizers, or moving lights. This work uses a light field camera, which records both where light lands and the direction it arrives from. Each sub-aperture view is dehazed in 2D, the results are merged into a smoothed central view, and that view guides filtering of the full light field. Tests used real data from a lab water tank, with in-air reference images, and the dataset is public.
Abstract
Underwater vision is subject to effects of underwater light propagation that act to absorb, scatter, and attenuate light rays between the scene and the imaging platform. Backscattering has been shown to have a strong impact on underwater images. As light interacts with particulate matter in the water column, it is scattered back towards the imaging sensor, resulting in a hazy effect across the image. A similar effect occurs in terrestrial applications in images of foggy scenes due to interaction with the atmosphere. Prior work on multi-image dehazing has relied on multiple cameras, polarization filters, or moving light sources. Single image dehazing is an ill-posed problem; proposed solutions rely on strong priors of the scene. This paper presents a novel method for underwater image dehazing using a light field camera to capture both the spatial and angular distribution of light across a scene. First, a 2D dehazing method is applied to each sub-aperture image. These dehazed images are then combined to produce a smoothed central view. Lastly, the smoothed central view is used as a reference to perform guided image filtering, resulting in a 4D dehazed underwater light field image. The developed method is validated on real light field data collected in a controlled in-lab water tank, with images taken in air for reference. This dataset is made publicly available.