Halo Separation-guided Underwater Multi-scale Image Restoration

2026-05-11

AI Summary

This work addresses the challenge of foreground halos induced by artificial lighting in underwater imaging, which severely degrade image quality and are inadequately handled by existing enhancement methods. The study introduces, for the first time, a halo separation mechanism into underwater image restoration and proposes an iterative single-image correction framework comprising two subnetworks: a halo layer separation module and a multi-scale restoration module. The former leverages radial gradient constraints to achieve gradient-minimized halo estimation, while the latter jointly recovers occluded details. Trained on the UIEB and EUVP synthetic datasets and validated on real-world scenes with artificial illumination, the method effectively suppresses halo artifacts, significantly enhances visual quality, and improves performance in downstream vision tasks.

Abstract

Underwater images captured by Autonomous Underwater Vehicles (AUVs) are inevitably affected by artificial light sources, which often produce halos in the foreground of the camera and seriously interfere with the quality of the image. The existing underwater image enhancement methods fail to fully consider this key problem, and the robustness of processing images under artificial light scenes is poor. In practical applications, since underwater image enhancement itself is a very challenging task, the influence of artificial light sources will lead to serious degradation of image performance and affect subsequent vision tasks. In order to effectively deal with this problem, this paper designs a single halo image correction method based on an iterative structure. The network is mainly divided into two sub-networks, one is the halo layer separation sub-network which aims to separate the halo by gradient minimization, and the other is the multi-scale recovery sub-network which aims to recover the image information masked by halo. The UIEB and EUVP synthetic datasets are used for training to ensure that the network can fully learn the characteristics and laws of underwater halo images. Then a large number of halo images taken in an underwater environment with real artificial light are collected for testing. In addition, the brightness distribution characteristics of underwater halo images are analyzed and the radial gradient is introduced to constraint eliminate halo to improve the effect of underwater image restoration.

Problem

• underwater image restoration
• halo artifact
• artificial light interference
• image degradation
• vision tasks

Innovation

• halo separation
• underwater image restoration
• radial gradient constraint
• multi-scale recovery
• artificial light interference