中国光伏电站空间范围数据集(2010-2022年)
收藏资源简介:
研究团队提出了一个利用Sentinel-2和Landsat数据,结合深度学习和变化检测技术,提取光伏发电站空间范围和安装日期的框架。使用了先进的语义分割模型TransUNet来提取光伏发电站,并采用连续变化检测和分类(CCDC)算法估计每个光伏发站的安装日期。最终得到2010-2020年中国光伏电站空间范围数据集,主要范围为中国大陆,属性包含光伏电站安装时间,面积和占用土地的主要类型。
The research team proposed a framework that leverages Sentinel-2 and Landsat data, combined with deep learning and change detection techniques, to extract the spatial extent and installation dates of photovoltaic (PV) power stations. The advanced semantic segmentation model TransUNet was employed to extract PV power stations, while the Continuous Change Detection and Classification (CCDC) algorithm was adopted to estimate the installation date of each PV power station. Ultimately, a spatial extent dataset of China's PV power stations spanning from 2010 to 2020 was developed, which primarily covers mainland China. The dataset includes attributes such as the installation time, area, and main land use type occupied by the PV power stations.




