地理本体驱动的面向对象滑坡识别  

Geographic Ontology-driven Object oriented Landslide Recognition

作　　者：魏家旺 惠文华[1] 程梦真 李海[1] WEI Jiawang;HUI Wenhua;CHENG Mengzhen;LI Hai(College of Geological Engineering and Geomatics,Chang’an University,Xi’an 710064,China)

机构地区：[1]长安大学地质工程与测绘学院,西安710064

出　　处：《遥感信息》2020年第2期94-99,共6页Remote Sensing Information

基　　金：国家自然科学基金项目(41731066)。

摘　　要：高空间分辨率遥感影像的出现促进了面向对象影像分析技术的发展,面向对象影像分析技术也由此应用到滑坡信息提取,但现阶段因为该技术对专家知识利用不够,导致其仍不能脱离传统的数据驱动的模式识别范畴,专家分析不同带来对象分类结果不一致,自动化程度较低。针对上述问题,引入知识工程中本体概念,对滑坡地理本体进行表达、建模,实现滑坡信息的自动提取。以甘肃黑方台地区的高分遥感影像数据为例,以地理本体驱动的解译框架进行了滑坡识别实验。实验表明,基于本体概念进行面向滑坡地理对象的信息提取是一种较好的思路,有利于进一步提高滑坡信息计算机提取的自动化程度和精度。The emergence of high spatial resolution remote sensing images has promoted the development of objectoriented image analysis technology,and object-oriented image analysis technology has also been applied to the landslide information extraction. However,because of insufficient use of expert knowledge at this stage,the object-oriented classification technology still cannot be separated from traditional data-driven pattern recognition,different expert analysis results in different object classification,and degree of automation is very low.Aiming at the above problems,the ontology concept in knowledge engineering is introduced to express and model the landslide geographic ontology,so as to realize the automatic extraction of landslide information.This paper uses the geographic ontology-driven interpretation framework to carry out geographic ontology-driven landslide recognition experiments in Heifangtai,Gansu Province.Experimental results show that the object-oriented landslide recognition based on ontology concept can effectively extract landslide information.

关 键 词：面向对象 滑坡识别 滑坡地理本体 高分影像 语义分类 

分 类 号：P237[天文地球—摄影测量与遥感]