An Underwater Dehazing Network with Implicit Transmission Estimation

TL;DR

The paper introduces UDehaze-iT, a deep learning model for underwater image enhancement that estimates scene depth and transmission implicitly, improving visual quality in underwater images.

Contribution

It proposes a novel underwater dehazing network that combines physical modeling with learnable parameters, achieving better interpretability and performance.

Findings

Achieves competitive results on UIEB and UFO-120 datasets.

Uses a composite loss function for effective training.

Estimates scene depth and transmission implicitly with learnable parameters.

Abstract

Underwater images suffer from wavelength-dependent light absorption and scattering, which reduces visual quality. This phenomenon could limit the operational reliability of autonomous underwater vehicles, marine surveys, and offshore inspection systems. Purely classical methods often achieve suboptimal performance in real-world datasets, while purely data-driven methods lack physical interpretability. In this letter, we propose UDehaze-iT, a deep network for underwater image enhancement that estimates scene depth implicitly and derives per-channel transmission through the Beer-Lambert law with learnable attenuation coefficients. We estimate atmospheric light as a semi-classical per-channel scalar, and a zero-initialized residual refiner corrects remaining artefacts after dehazing. To effectively train our method, we apply a composite loss function consisting of five key terms: a L1 loss, a multi-scale patchwise DCT loss, a forward model reconstruction loss, and two regularization terms. With ~0.9M parameters, UDehaze-iT achieves competitive performance on UIEB and UFO-120 datasets.