Marine Snow Removal Benchmarking Dataset

TL;DR

The paper addresses the lack of ground-truth data for marine snow removal in underwater imagery by introducing the Marine Snow Removal Benchmarking Dataset (MSRB). It defines two representative artifact models, Highland Type and Volcanic Crater Type, and demonstrates a synthesis pipeline to generate paired ground-truth and degraded images for two MSR tasks. It provides the first benchmarking of MSR methods, showing that a deep network (U-Net) can outperform median-filter baselines on synthetic data and has some success on real images with limitations. The dataset is publicly available and is intended to accelerate development of robust MSR methods and facilitate objective evaluation.

Abstract

This paper introduces a new benchmarking dataset for marine snow removal of underwater images. Marine snow is one of the main degradation sources of underwater images that are caused by small particles, e.g., organic matter and sand, between the underwater scene and photosensors. We mathematically model two typical types of marine snow from the observations of real underwater images. The modeled artifacts are synthesized with underwater images to construct large-scale pairs of ground truth and degraded images to calculate objective qualities for marine snow removal and to train a deep neural network. We propose two marine snow removal tasks using the dataset and show the first benchmarking results of marine snow removal. The Marine Snow Removal Benchmarking Dataset is publicly available online.

Figures (9)

• Figure 1: Marine snow artifacts in real underwater images.
• Figure 2: Marine snow examples in real underwater images. Top: Highland type (type H). Bottom: Volcanic Crater type (type V). The 3D plots correspond to the grayscale images.
• Figure 3: Ellipses used for marine snow synthesis.
• Figure 4: Proposed marine snow models. Top: Type H marine snow. Bottom: Type V marine snow.
• Figure 9: Limitations. Top: Original images. Bottom: Restoration results by U-Net.