monster-monash/S2Agri-17

S2Agri-17是一个卫星图像时间序列分类数据集，包含覆盖法国一个12,100平方公里区域的Sentinel-2卫星数据。该数据集包含10个光谱波段，时间序列长度为24，观测时间为2017年1月至10月。数据集经过重新组织，适用于基于像素的处理，共有59,268,823个像素。数据集提供了两组土地覆盖分类标签，分别为17个类别和34个类别。数据集基于法国地块识别系统进行分类，且是不平衡的，最大的四个类别占据了90%的地块。数据集分为两个版本：S2Agri-17和S2Agri-34，以及两个较小的随机10%地块数据版本：S2Agri-10pc-17和S2Agri-10pc-34。

S2Agri-17 is a satellite image time series classification dataset containing Sentinel-2 satellite data over a 12,100 km² area in France. The dataset includes 10 spectral bands with time series of length 24, observed from January to October 2017. The dataset has been reorganized for pixel-based processing and consists of 59,268,823 pixels. Two sets of land cover classification labels are provided, with 17 and 34 classes respectively. The dataset is classified based on the French Land Parcel Identification System and is imbalanced, with the largest four classes accounting for 90% of the parcels. The dataset is available in two versions: S2Agri-17 and S2Agri-34, as well as two smaller versions with data for 10% of the land parcels randomly selected: S2Agri-10pc-17 and S2Agri-10pc-34.