Sentinel2-image-sharpening-dataset
收藏资源简介:
我们提供了一个数据集,用于评估Sentinel-2图像锐化模型。该数据集包含17个Sentinel-2场景,其获取日期和瓦片位置在下面的论文中的表2中列出。数据集包括Sentinel-2的10/20/60米波段。这些场景是在2021年跨越中国的四个季节收集的。不同的空间位置、海拔条件和获取日期确保了收集数据的土地覆盖类型和时空特性的多样性,基于此,可以充分测试图像锐化模型。值得注意的是,在具有复杂土地覆盖的异质景观上进行图像锐化任务始终被认为是具有挑战性的。因此,我们特别关注异质景观,如碎片农田、湿地地区和城市区域,以揭示模型在具有挑战性的场景中的锐化性能。
We provide a dataset for evaluating Sentinel-2 image sharpening models. This dataset comprises 17 Sentinel-2 scenes, with their acquisition dates and tile locations listed in Table 2 of the referenced paper. The dataset includes Sentinel-2's 10/20/60-meter bands. These scenes were collected across four seasons in China during 2021. The diversity in spatial locations, elevation conditions, and acquisition dates ensures a variety of land cover types and spatiotemporal characteristics in the collected data, based on which the image sharpening models can be thoroughly tested. It is noteworthy that image sharpening tasks on heterogeneous landscapes with complex land covers are always considered challenging. Therefore, we pay special attention to heterogeneous landscapes such as fragmented farmland, wetland areas, and urban regions to reveal the sharpening performance of models in challenging scenarios.
Sentinel2-image-sharpening-dataset 概述
数据集内容
- 场景数量:包含17个Sentinel-2场景。
- 数据细节:每个场景包括10/20/60米的波段数据。
- 采集时间与地点:数据采集于2021年,覆盖中国的四个季节,具体采集日期和地点详见论文中的Table 2。
- 地理多样性:数据集涵盖不同空间位置、海拔条件和采集日期,确保了土地覆盖类型和时空特性的多样性。
数据集用途
- 评估目的:用于评估Sentinel-2图像锐化模型。
- 挑战性场景:特别关注异质景观,如零散农田、湿地和城市区域,以测试模型在复杂环境下的锐化性能。
数据集获取
- 下载链接:数据集可通过以下链接获取:链接
- 提取码:kqpa
引用信息
- 参考文献:Wu, J., Lin, L., Zhang C., Li T., Cheng, X., Nan, F., 2022, Generating Sentinel-2 all-band 10-m data by sharpening 20/60-m bands: a hierarchical fusion network, ISPRS Journal of Photogrammetry and Remote Sensing, Accepted.




