Perceptual Underwater Image Enhancement With Deep Learning and Physical Priors

TLDR

Underwater image enhancement is crucial for object detection in navigation and exploration, yet most existing methods treat enhancement and detection as separate modules, which can hinder detection performance. This study introduces two perceptual enhancement models that jointly integrate a deep enhancement network with a detection perceptor. The detection perceptor supplies gradient feedback to steer the enhancement network toward visually pleasing and detection‑friendly outputs, while a hybrid synthesis model combining physical priors and data‑driven cues generates training data for real‑world scenarios. Experiments demonstrate that the proposed approach surpasses several state‑of‑the‑art methods on both real‑world and synthetic underwater datasets.

Abstract

Underwater image enhancement, as a pre-processing step to support the following object detection task, has drawn considerable attention in the field of underwater navigation and ocean exploration. However, most of the existing underwater image enhancement strategies tend to consider enhancement and detection as two fully independent modules with no interaction, and the practice of separate optimisation does not always help the following object detection task. In this article, we propose two perceptual enhancement models, each of which uses a deep enhancement model with a detection perceptor. The detection perceptor provides feedback information in the form of gradients to guide the enhancement model to generate patch level visually pleasing or detection favourable images. In addition, due to the lack of training data, a hybrid underwater image synthesis model, which fuses physical priors and data-driven cues, is proposed to synthesise training data and generalise our enhancement model for real-world underwater images. Experimental results show the superiority of our proposed method over several state-of-the-art methods on both real-world and synthetic underwater datasets.