摘要
针对耕地“非农化”行为的快速发现与定位需求,通过综合分析深度学习算法在卫星遥感地物检测和变化发现中的应用案例,提出了一种基于深度学习的耕地“非农化”快速遥感监测技术路线。选取实验区,采用基于DeepLabV3+语义分割的深度学习遥感影像地物变化检测方法对典型耕地“非农化”行为进行了变化图斑监测自动提取与后处理,并验证了技术路线的可行性,提高了耕地“非农化”疑问图斑提取的效率。
In order to meet the needs of rapid detection of non-agricultural cultivated land use, we proposed a technical design of rapid cultivated land remote sensing monitoring based on deep learning by comprehensively analyzing the cases of deep learning algorithm in satellite remote sensing object detection and change detection. In the experiment area, we used the deep learning surface feature change detection method based on DeepLab V3+ semantic segmentation to detect and post process some typical non-agricultural cultivated land use automatically, which could verify the technical design feasibility and improve non-agricultural cultivated land detection efficiency.
作者
石婷婷
戴腾
厉芳婷
王爱华
SHI Tingting;DAI Teng;LI Fangting;WANG Aihua(Hubei Institute of Photogrammetry and Remote Sensing,Wuhan 430074,China;Hubei Institute of Surveying and Mapping Engineering,Wuhan 430074,China;Shandong Provincial Institute of Land Surveying and Mapping,Jinan 250013,China)
出处
《地理空间信息》
2022年第3期34-37,共4页
Geospatial Information
基金
湖北省自然资源厅科技资助项目(ZRZY2020KJ09)。
关键词
耕地“非农化”
深度学习
遥感监测
non-agricultural cultivated land use
deep learning
remote sensing monitoring
作者简介
第一作者:石婷婷(1986—),工程师,主要研究方向为地理信息相关技术与应用。
