Remote Sensing Imagery for Flood Detection: Exploration of Augmentation Strategies

TL;DR

This paper tackles flood detection from remote sensing imagery by applying a comprehensive augmentation-ablation study on high-resolution RGB river images (BlessemFlood21) to improve semantic water segmentation. It systematically groups augmentations into nine categories using Albumentations and evaluates two state-of-the-art segmentation models (UNet++ and DeepLabV3+) across 100, 200, and 300 training epochs. Key findings show distortion-based augmentations and blur/noise augmentations provide the largest accuracy gains, with longer training further boosting performance and reducing cross-group variance. The work provides a structured framework for incorporating augmentation strategies in flood mapping and informs practical training choices for high-resolution RS-based flood detection.

Abstract

Floods cause serious problems around the world. Responding quickly and effectively requires accurate and timely information about the affected areas. The effective use of Remote Sensing images for accurate flood detection requires specific detection methods. Typically, Deep Neural Networks are employed, which are trained on specific datasets. For the purpose of river flood detection in RGB imagery, we use the BlessemFlood21 dataset. We here explore the use of different augmentation strategies, ranging from basic approaches to more complex techniques, including optical distortion. By identifying effective strategies, we aim to refine the training process of state-of-the-art Deep Learning segmentation networks.