Flood Mapping from RGB imagery using a Vision Foundation Model

📄 arXiv: 2606.24120v1 📥 PDF

作者: Vladyslav Polushko, Tilman Bucher, Ronald Rösch, Thomas März, Markus Rauhut, Andreas Weinmann

分类: cs.CV, eess.IV

发布日期: 2026-06-23

💡 一句话要点

提出基于视觉基础模型的RGB图像洪水映射方法

🎯 匹配领域: 支柱九：具身大模型 (Embodied Foundation Models)

关键词: 洪水映射 视觉基础模型 深度学习 迁移学习 水体分割 RGB图像 卫星数据 应急响应

📋 核心要点

  1. 现有的基于CNN和小型视觉变换器的洪水映射方法需要大量数据以适应不同的洪水场景，限制了其应用。
  2. 本文提出了一种新的模型Prithvi-2.0-UPN，通过微调卫星预训练的视觉基础模型，旨在提高RGB图像洪水映射的准确性。
  3. 实验结果显示，Prithvi-2.0-UPN在新洪水事件的迁移学习中表现优异，尤其在使用少量训练数据时，性能提升显著。

📝 摘要（中文）

及时、高分辨率的洪水范围地图对于应急响应和损失评估至关重要。本文考虑使用低成本的空中RGB图像进行洪水映射。尽管深度学习模型在水体分割中表现良好，但它们需要大量数据以适应不同的洪水事件。我们研究了如何将预训练的卫星基础模型适应于厘米级的RGB图像洪水映射，提出了Prithvi-2.0-UPN模型，并在两个RGB数据集上进行了微调。实验结果表明，该模型在BlessemFlood21和NeuenahrFlood数据集上达到了最先进的结果，并在零样本设置下优于现有基线模型，显示出良好的迁移能力。

🔬 方法详解

问题定义：本文旨在解决现有洪水映射方法在不同洪水事件中适应性差的问题，尤其是基于CNN和小型视觉变换器的模型需要大量数据进行训练。

核心思路：通过微调一个预训练的卫星基础模型Prithvi-2.0-UPN，使其能够适应RGB图像的洪水映射任务，从而提高模型的泛化能力和准确性。

技术框架：整体架构包括一个视觉变换器Prithvi-EO-2.0-600M和一个UPerNet解码器，专注于二元水体分割。模型在两个RGB数据集（BlessemFlood21和NeuenahrFlood）上进行微调和评估。

关键创新：最重要的创新在于将卫星预训练模型应用于RGB图像洪水映射，展示了其在迁移学习中的有效性，尤其是在零样本设置下的优越性能。

关键设计：在模型训练中，采用了适当的损失函数和网络结构设计，确保模型能够快速适应新数据集，并通过小规模的训练数据进一步提升性能。

🖼️ 关键图片

fig_0
fig_1
fig_2

📊 实验亮点

在实验中，Prithvi-2.0-UPN在BlessemFlood21和NeuenahrFlood数据集上达到了最先进的结果，并在零样本设置下的迁移学习中优于现有基线模型，显示出显著的性能提升，尤其是在使用少量训练数据时，模型的表现几乎达到了完全训练的水平。

🎯 应用场景

该研究的潜在应用领域包括自然灾害管理、城市规划和环境监测。通过提供高分辨率的洪水映射，能够帮助应急响应团队快速评估灾害影响，制定有效的救援策略，进而减少损失和提高救援效率。未来，该方法还可扩展到其他类型的自然灾害监测和评估中。

📄 摘要（原文）

Timely, high-resolution maps of flood extent around settlements are essential for emergency response and damage assessment. We consider airborne RGB imagery for flood mapping as it can be collected rapidly at low cost. To produce flood maps, deep learning models for water segmentation are often used. CNN based and small vision transformer models are used. However, they need much data for adaptation to a change of scenery, i.e., another flooding event. Vision foundation models or large vision transformers are known to generalize across domains. Recently, foundation models for Earth observation became available. They are pretrained on satellite data, whose spatial resolution, viewing geometry, and radiometry differ from nadir RGB imagery. Thus, adaptation is required. We investigate how a satellite-pretrained Earth observation foundation model can be adapted to centimeter-scale floodwater mapping from RGB imagery. Specifically, we fine-tune a model we call Prithvi-2.0-UPN consisting of the Prithvi-EO-2.0-600M Vision Transformer combined with a UPerNet decoder for binary water segmentation on two RGB datasets (BlessemFlood21, NeuenahrFlood). In a first experiment we observe that Prithvi-2.0-UPN reaches state-of-the-art results on BlessemFlood21 and NeuenahrFlood, when trained on their datasets. In a second experiment we show that Prithvi-2.0-UPN performs better than state-of-the-art baseline models for transfer to a new flood event (trained on BlessemFlood21, tested on NeuenahrFlood) in a zero-shot setting. However, the performance indicates room for improvement. In this respect, we investigate in a third experiment how performance improves when further fine-tuning the models with small shares of NeuenahrFlood training data: Prithvi-2.0-UPN improves the fastest and reaches almost the performance level when fully trained on NeuenahrFlood, indicating transfer capabilities.