2015年中国长三角地区GF1-2m土地利用分类数据集

基础影像数据以2015年长三角地区GF1夏季影像为主，缺失部分影像用2015年前后的夏季GF2及GF1影像补充。预处理工作包括影像适度去云及图像增强。
利用人工勾绘的分类数据和遥感影像制作训练样本。综合考虑样本均衡性、代表性、分布合理等因素。利用深度学习技术，进行模型训练及分类处理，制作的长三角地区2015年2m分辨率土地利用分类数据集。

The foundational image dataset primarily consists of summer Gaofen-1 (GF1) imagery acquired over the Yangtze River Delta (YRD) in 2015. Missing imagery was supplemented with summer Gaofen-2 (GF2) and GF1 imagery collected around the year 2015. Preprocessing operations included appropriate cloud removal and image enhancement for the remote sensing data. Training samples were created using manually annotated classification data and the preprocessed remote sensing imagery, with comprehensive considerations given to factors including sample balance, representativeness, and reasonable spatial distribution. Leveraging deep learning techniques for model training and classification, a 2-meter resolution land use classification dataset of the Yangtze River Delta in 2015 was finally developed.