深度学习网络支持下的农房侵占耕地自动化监测

近年来,违法占用耕地现象屡禁不止,如何利用人工智能等新一代信息技术,快速摸清农村乱占耕地建房底数,做到"早发现、早制止、严查处",是当前整治农村乱占耕地建房工作的研究难点之一。本文通过对高分辨率自然资源影像数据进行预处理,构建基于深度学习网络的自动化监测模型,应用模型进行预测并对输出结果进行GIS优化和空间叠加。试验结果表明,该方法可以快速监测出疑似侵占耕地的违法房屋,为坚守"耕地红线不突破"的底线提供了智能化技术选择,可服务于整治农村乱占耕地建房工作。

Automatic monitoring of farmland occupation by farm house based on deep learning network

In recent years,illegal occupation of arable land has been repeatedly prohibited.How to use artificial intelligence and other new-generation information technology to quickly figure out the number of illegal occupation of arable land and build houses in rural areas,and achieve"early detection,early stop,strict investigation and punishment ",is one of the current research difficulties in the work of rectifying the illegal occupation of farmland in rural areas.This paper preprocesses high-resolution natural resource image data,and then builds an automated monitoring model based on a deep learning network.Thirdly,it applies model to predict and GIS optimization and spatial overlay of output results.Experimental results show that this method can quickly detect illegal houses that are suspected of occupying cultivated land,and provides intelligent technology options for sticking to the bottom line of" not breaking through the red line of cultivated land",and can serve the work of rectifying building houses on the cultivated land in rural areas.