Sentinel Flooding Sequences Dataset
收藏资源简介:
该数据集包含ESA的Sentinel卫星的多时相卫星图像补丁,用于洪水事件检测。数据集最初作为MediaEval基准测试2019年多媒体卫星任务的一个共享子任务发布。参与者接收一系列Sentinel-2卫星图像序列,需要创建一个二元分类器来确定该区域是否正在发生洪水事件。
This dataset comprises multi-temporal satellite image patches from ESA's Sentinel satellites, designed for flood event detection. Initially released as a shared subtask for the MediaEval Benchmark 2019 Multimedia Satellite Task, participants were provided with a series of Sentinel-2 satellite image sequences. The objective was to develop a binary classifier to ascertain whether a flood event was occurring in the specified area.
数据集概述
数据集名称
Sentinel Flooding Sequences Dataset
数据集内容
该数据集包含由ESA的Sentinel卫星获取的多时相卫星图像补丁,用于洪水事件检测。数据集包括412个卫星序列,涵盖2018/2019年的洪水和非洪水事件。
数据集结构
- 类标签:每个图像序列都附带一个类标签,用于指示是否存在洪水事件。
- 图像波段时间序列:每个图像序列包含一个时间序列文件(timeseries.txt),记录每个时间点的日期、洪水状态和全数据覆盖情况。
数据集下载
数据集更新计划
未来将持续更新数据集,包括更多近期洪水事件,以评估方法对不同事件和地理区域的泛化能力,并计划提供额外的Ground Truth和其他模态数据。
引用信息
@inproceedings{bischke_2019_ceurws_mmsat19, title = {The Multimedia Satellite Task at MediaEval 2019: Emergency Response for Flooding Events}, author = {Bischke, Benjamin and Helber, Patrick and Brugman, Simon and Basar, Erkan and Zhao, Zhengyu and Larson, Martha and Pogorelov, Konstantin}, booktitle = {Working Notes Proceedings of the MediaEval 2019 Workshop}, year = {2019}, address={Sophia Antipolis, France}, organization={CEUR-WS}, }




