面向对象灾害遥感判读知识库设计与实现  

Design and Implementation of Disasters Knowledge Base for Remote Sensing Images Interpretation

作　　者：任玉环[1] 刘亚岚[1] 

机构地区：[1]中国科学院遥感与数字地球研究所,北京100101

出　　处：《测绘与空间地理信息》2016年第7期5-9,14,共6页Geomatics & Spatial Information Technology

基　　金：北京市自然科学基金项目(4144088);国家自然科学基金重点基金预研项目(91024004);国家自然基金面上项目(40971201)资助

摘　　要：灾害遥感研究与应用中积累的判读知识是灾害应急中灾情信息快速、准确提取的重要依据,为了重复利用专家所掌握的不可多得的判读知识,需要通过知识库将其存储管理起来,以实现共享和利用。本文采用面向对象和模糊加权产生式规则的知识表达方法,研制了灾害遥感判读知识库,实现了知识获取、管理与推理,为判读人员提供了学习灾害判读知识和灾害应急应用的平台,对自动化和智能化的灾害遥感信息提取研究将起到推动作用。Interpretation knowledge is the basis of disaster information extraction from remote sensing images quickly and accurately. Hence, interpretation knowledge is vital for disaster emergency response. However, experts usually accumulate interpretation knowledge during research and applications of disaster remote sensing, and others are difficult to reuse it. This paper proposed the interpretation knowledge framework in remote sensing images firstly. Then, it designed a knowledge representation approach based on object - oriented method and weighted fuzzy production rules. Finally, it implemented the Disasters Knowledge Base for Remotely Sensed Images Interpretation （DKB/RSII） for management of quantitative and qualitative knowledge. DKB/RSII can manage plenty of implication knowledge intensively via knowledge acquisition, representation and storage. Meanwhile, DKB/RSII can help to query and share the disasters knowledge. The result will promote the researches of disaster information extraction from remote sensing images automati- cally and intelligently.

关 键 词：灾害 遥感 判读 影像特征 知识库 规则 

分 类 号：P208[天文地球—地图制图学与地理信息工程]