面向对象的遥感分类方法在输电线路地表信息提取中的应用

面向对象的遥感分类方法是影像分辨率越来越高的背景下的产物。本文以某特高压输电线路工程为研究对象,选取宁夏、陕西、山西、河南、安徽境内五个典型地区样本,通过影像特征分析、多尺度分割和地物分类建立了水体、建筑、林地、道路等四种典型地物的分类规则集,并对工程全线进行信息提取和精度分析。结果表明:利用面向对象的方法提取高分辨率航空影像地表信息能够为输电线路智能选线快速提供基础地理空间数据。

Land-cover Classification of Electric Transmission Lines of Hierarchical Classification Based on Object-Oriented

Object-oriented remote sensing classification method is the product of the background of higher and higher image resolution.Taking an UHV transmission line project as the research object,this paper selects five typical areas in Ningxia,Shaanxi,Shanxi,Henan and Anhui as samples,establishes the classification rule set of four typical land objects,including water,buildings,woodlands and roads,through image feature analysis,multi-scale segmentation and classification.Information extraction and accuracy analysis were carried out.The results show that extracting high resolution aerial image surface information by object-oriented method can provide basic geospatial data for intelligent line selection.