PV power plants of China from 2010 to 2022

We develops a framework to extract the spatial extent and installation date of PV power plants from Sentinel-2 and Landsat data using deep learning and change detection techniques in China. A geospatial dataset (Shapefile format) of PV polygons with installation dates in China from 2010 to 2022 is obtained with the F1-score of 96.08% for its spatial extent and the overall accuracy of 89.86% for its installation dates. The dataset is associated with the paper of "Yuehong Chen, Jiayue Zhou, Yong Ge and Jinwei Dong. Uncovering the rapid expansion of photovoltaic power plants in China from 2010 to 2022 using satellite data and deep learning, 2024" published in the journal of Remote Sensing of Environment.

本研究开发了一套框架，采用深度学习与变化检测技术，从哨兵-2（Sentinel-2）和陆地卫星（Landsat）遥感数据中提取中国境内光伏电站（PV power plants）的空间范围与安装日期。最终构建得到2010至2022年中国境内带安装日期的光伏多边形地理空间数据集（格式为Shapefile），该数据集的空间范围识别F1分数（F1-score）达96.08%，安装日期标注的总体精度为89.86%。该数据集关联发表于《环境遥感（Remote Sensing of Environment）》期刊的论文"Yuehong Chen, Jiayue Zhou, Yong Ge and Jinwei Dong. Uncovering the rapid expansion of photovoltaic power plants in China from 2010 to 2022 using satellite data and deep learning, 2024"。

搜集汇总

背景与挑战

背景概述

该数据集包含中国2010-2022年光伏电站的地理空间数据(Shapefile格式)，通过卫星遥感和深度学习技术获取，具有高精度的空间范围识别(96.08% F1-score)和安装日期判定(89.86%准确率)，相关研究成果发表在《Remote Sensing of Environment》期刊。