arXiv 2602.126522026-02cs.CV

CBEN -- 一种用于云鲁棒遥感图像理解的多模态机器学习数据集

CBEN -- A Multimodal Machine Learning Dataset for Cloud Robust Remote Sensing Image Understanding

Marco Stricker, Masakazu Iwamura, Koichi Kise

备注：: We are currently in the process of selecting an appropriate journal for submission

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文提出CBEN数据集，用于研究云干扰对遥感图像理解的影响，通过结合光学与雷达数据，验证云鲁棒方法在云覆盖场景下的有效性。

摘要：

云是光学卫星影像中常见的现象，会干扰遥感图像，给遥感带来挑战。然而，在文献中，云无分析经常被采用，即排除云图像的机器学习数据集和方法。这种做法无法应用于时间敏感的应用，例如自然灾害期间。可能的解决方案是在预处理步骤中应用云去除，以确保在这些条件下云自由解决方案不会失败。但云去除方法仍在积极研究中，存在诸如生成视觉伪影等缺点。因此，开发对云天气影响较小的云鲁棒方法是有必要的。云鲁棒方法可以通过结合光学数据与雷达数据实现，雷达数据不受云影响。虽然许多用于机器学习的数据集结合了光学和雷达数据，但大多数研究人员排除了云图像。我们发现从机器学习训练和评估中排除云图像是一种限制，减少了在云覆盖场景中的适用性。为此，我们组建了一个数据集，名为CloudyBigEarthNet（CBEN），包含配对的光学和雷达图像，用于训练和评估。使用平均精度（AP）作为评估指标，我们显示，基于结合的晴空光学和雷达影像的最先进方法在评估云图像时性能下降了23-33个百分点。然后，我们适应这些方法在训练过程中使用云光学数据，相对于原始方法，在云测试案例上实现了17.2-28.7个百分点的相对提升。代码和数据集可在：https://github.com/mstricker13/CBEN 公开获取。

英文摘要：

Clouds are a common phenomenon that distorts optical satellite imagery, which poses a challenge for remote sensing. However, in the literature cloudless analysis is often performed where cloudy images are excluded from machine learning datasets and methods. Such an approach cannot be applied to time sensitive applications, e.g., during natural disasters. A possible solution is to apply cloud removal as a preprocessing step to ensure that cloudfree solutions are not failing under such conditions. But cloud removal methods are still actively researched and suffer from drawbacks, such as generated visual artifacts. Therefore, it is desirable to develop cloud robust methods that are less affected by cloudy weather. Cloud robust methods can be achieved by combining optical data with radar, a modality unaffected by clouds. While many datasets for machine learning combine optical and radar data, most researchers exclude cloudy images. We identify this exclusion from machine learning training and evaluation as a limitation that reduces applicability to cloudy scenarios. To investigate this, we assembled a dataset, named CloudyBigEarthNet (CBEN), of paired optical and radar images with cloud occlusion for training and evaluation. Using average precision (AP) as the evaluation metric, we show that state-of-the-art methods trained on combined clear-sky optical and radar imagery suffer performance drops of 23-33 percentage points when evaluated on cloudy images. We then adapt these methods to cloudy optical data during training, achieving relative improvement of 17.2-28.7 percentage points on cloudy test cases compared with the original approaches. Code and dataset are publicly available at: https://github.com/mstricker13/CBEN

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。

AI总结：

摘要：

英文摘要：

相关论文