Urban vacant land of 36 major Chinese cities

The data include two parts: (1) urban vacant land of 36 major Chinese cities (in shapefile format), and (2) codes and data used for automatic vacant land identification.<br>Urban vacant land is a growing issue worldwide. The study is aimed to realize large-scale automatic identification of urban vacant land. A framework based on deep learning techniques is proposed to extract urban vacant land of 36 major Chinese cities through semantic segmentation of high-resolution remote sensing images. The automatic identification framework is proved to be accurate and efficient, with strong robustness. This method is expected to serve as a practical approach in various countries and regions. The data of urban vacant land of 36 cities can be used in further studies.