城市建筑2026，Vol.23Issue(17)：124-126,3.DOI:10.19892/j.cnki.csjz.2026.17.29
基于深度学习时序制图的夷陵区国土空间规划土地利用研究
Research on Land Use in Territorial Spatial Planning of Yiling District Based on Deep Learning Time-Series Mapping
摘要
Abstract
To address the problems of low resolution and long update cycle of the traditional land use data in the complex mountainous areas of the Three Gorges Reservoir Area,which can hardly support the refined management and control of the"Three Zones and Three Lines",this paper takes Yiling District in Yichang City as the research area,and conducts land use mapping by using the TSSCD deep learning model based on 10 m resolution Sentinel-2 time-series images on GEE platform.After localized parameter tuning,the classification Kappa coefficient of the model reaches 0.884.The analysis on high-precision data from 2019 to 2022 shows that during the study period,the net expansion of impervious surface reached 54.47 km2,and the net reduction of cultivated land reached 73.69 km2,with prominent spatial conflicts among urbanization,ecological protection and cultivated land protection.The research results can provide support for the refined territorial spatial governance of mountainous counties.关键词
三区三线/时序土地利用制图/国土空间规划/夷陵区Key words
Three Zones and Three Lines/time-series land use mapping/territorial spatial planning/Yiling District分类
建筑与水利引用本文复制引用
张攀,明涛..基于深度学习时序制图的夷陵区国土空间规划土地利用研究[J].城市建筑,2026,23(17):124-126,3.基金项目
国家民用空间基础设施陆地观测卫星共性应用支撑平台项目"十三五空基真实性检验站(宜昌站)建设"(2017-00052-73-01-001735) （宜昌站）
