CCESAR: Coastline Classification-Extraction From SAR Images Using CNN-U-Net Combination

TL;DR

This paper introduces CCESAR, a two-stage deep learning approach combining classification and segmentation to improve coastline extraction from SAR images, outperforming single-model methods.

Contribution

The paper presents a novel two-stage CNN-U-Net based workflow for coastline classification and extraction from SAR images, addressing limitations of single-model segmentation.

Findings

Two-stage workflow outperforms single U-Net model

Effective coastline classification across different image compression levels

Improved accuracy in coastline extraction from Sentinel-1 SAR images

Abstract

In this article, we improve the deep learning solution for coastline extraction from Synthetic Aperture Radar (SAR) images by proposing a two-stage model involving image classification followed by segmentation. We hypothesize that a single segmentation model usually used for coastline detection is insufficient to characterize different coastline types. We demonstrate that the need for a two-stage workflow prevails through different compression levels of these images. Our results from experiments using a combination of CNN and U-Net models on Sentinel-1 images show that the two-stage workflow, coastline classification-extraction from SAR images (CCESAR) outperforms a single U-Net segmentation model.