dm4eo/ssl4eo_eu_forest

SSL4EO-EU-Forest数据集是一个包含Sentinel-2影像和欧洲森林遮罩的集合，用于森林土地覆盖的二进制语义分割。该数据集包括16K个Sentinel-2 L2A影像的空间块，具有12个波段和10米像素分辨率，涵盖了2018年的四个季节。遮罩是从2018年HRL数据集派生出来的，将区域分类为无树覆盖区域、阔叶林和针叶林。数据按地理位置和季节组织在子目录中。该数据集是SSL4EO项目的一部分，并由欧盟的地平线欧洲计划资助。

The SSL4EO-EU-Forest dataset is a collection of Sentinel-2 imagery and forest masks for European forests, designed for binary semantic segmentation of forest land cover. It consists of 16K spatial patches of Sentinel-2 L2A imagery with 12 bands and 10m pixel resolution, covering the four seasons of the year 2018. The masks are derived from the 2018 HRL dataset and classify areas into non-tree covered, broadleaved forest, and coniferous forest. The data is organized in subdirectories by geolocation and season. This dataset is part of the SSL4EO project and is funded by the European Unions Horizon Europe program.