摘要
基于深度学习进行遥感影像变化检测,对两幅不同时相的遥感影像进行校正、影像配准、生成变化检测图、隐患设施识别,最终计算燃气管道、输气站与隐患设施的隐患距离,判断隐患水平。分析基于遥感影像变化检测的燃气管道隐患排查流程。在隐患设施中(民房、围墙、铁塔、电杆、铁路、公路、厂矿企业等),除民房、围墙外,其他设施的修建需要有关部门的审批,因此安全隐患主要来自民房、围墙。
Remote sensing image change detection is carried out based on deep learning. The remote sensing images of the two non-simultaneous phases are corrected,the image registration,generation of change detection map,identification of hidden danger facilities are carried out,and the hidden danger distances between the gas pipelines,gas transmission stations and facilities are finally calculated,and hidden danger levels are judged. The hidden danger investigation process of gas pipeline is analyzed based on remote sensing image change detection. In hidden danger facilities( residential houses,fences,iron towers,electric poles,railways,highways,factories and mines and so on),except for residential houses and fences,the construction of other facilities requires the approval of the relevant departments. Therefore,the hidden dangers mainly come from the illegal residential houses and fences.
作者
焦建瑛
吴波
李夏喜
曹印锋
JIAO Jianying;WU Bo;LI Xiaxi;CAO Yinfeng
出处
《煤气与热力》
2019年第12期10025-10028,10045,10046,共6页
Gas & Heat
关键词
燃气管道隐患排查
遥感影像
变化检测
深度学习
hidden danger of gas pipeline
remote sensing image
change detection
deep learning
