高分2号影像在秦岭北麓西安段非农化建设用地遥感监测的应用  

Application of GF-2 image in remote sensing monitoring of nonagricultural construction land in Xi’an section of northern foothill of Qinling Mountains

作　　者：郑晨 赵燕伶 孙晨红 李紫涵 杨欢腾 ZHENG Chen;ZHAO Yanling;SUN Chenhong;LI Zihan;YANG Huanteng(Natural Resources Shaanxi Satellite Application Technology Center,Xi'an,Shaanxi 710054,China)

机构地区：[1]自然资源陕西省卫星应用技术中心,陕西西安710054

出　　处：《北京测绘》2025年第1期68-73,共6页Beijing Surveying and Mapping

基　　金：2023年度陕西省土地矿产卫片执法检查项目(YWGL-ZC-2023-0026)。

摘　　要：秦岭区域保护区众多,山前平原地区人类活动较多,为保证秦岭区域自然资源、山前平原地区耕地合理合规使用,减少区域范围内乱采乱挖、环境污染等现象,文章以高分2号影像为底图,采用基于全卷积神经网络(FCN)框架的深度学习图像分类方法,对秦岭北麓西安段区域非农建设用地信息进行提取。利用分类结果比较法进行变化信息提取,并分析变化原因。为后续国土部门土地利用决策科学化、拆除违建及复耕、复绿监管精准化以及土地资源公共服务便民化,提供强有力的技术支撑。There are many protected areas in the Qinling Mountains,and many human activities occur in the piedmont plain area.In order to ensure the rational and standard utilization of natural resources in the Qinling Mountains and the cultivated land in the piedmont plain area,reduce the phenomena of random excavation and environmental pollution in the area,and achieve cultivated land protection,this paper took GF-2 image as the base map and used the deep learning image classification method based on fully convolutional network(FCN)framework to extract the information of non-agricultural construction land in Xi’an section of the northern foothill of Qinling Mountains.The paper used the classification result comparison method to extract the change information and analyze the reasons for the change.It provided strong technical support for the scientific decision-making of follow-up land utilization by the national land sector,the demolition of illegal construction and re-cultivation,the precision of re-greening supervision,and the convenience of public services for land resources.

关 键 词：秦岭北麓西安段 非农建设用地 批量处理 自动解译 监测 

分 类 号：P237[天文地球—摄影测量与遥感]