基于高分影像的水利空间要素提取规则集构建  

Rule Sets Building of Water Conservancy Facilities Spatial Element Extraction Based on High Resolution Remote Sensing Image

作　　者：许莹[1] 张友静[1,2] 张琴 

机构地区：[1]河海大学地球科学与工程学院,江苏南京210098 [2]河海大学水文水资源与水利工程科学国家重点实验室,江苏南京210098 [3]南京物联网应用研究院有限公司,江苏南京210013

出　　处：《地理空间信息》2016年第5期71-74,7,共4页Geospatial Information

基　　金：高分辨率对地观测系统重大专项资助项目(08-Y30B07-9001-13/15)

摘　　要：针对高分遥感水利要素自动提取问题,提出了一种在融合处理后的高分一号(GF-1)遥感影像下表征和提取水利基础空间要素的方法。依据水利要素与水体的空间关系,采取先进行土地利用/覆盖分类,再提取水利设施要素的技术思路,构建了较为完整的水利空间要素提取规则集。利用混淆矩阵和矢量化图像对提取精度和叠合精度进行评价。结果表明,该规则集用于高分辨率影像水利基础空间要素提取的效果较好。其中,河流提取精度和叠合精度最高,分别为97.1%和94.1%;河流阻断物的提取精度和叠合精度最低,分别为86.0%和81.4%;其他水利要素总体精度较高。This paper provided a method of extracting the water conservancy facilities spatial element in GF-1 image after fusion processing based on the object-oriented approach.Research thought of this paper was extracting water conservancy facilities after the land cover classifications,which rely on the spatial relationship between water and conservancy facilities.And then,the paper built a complete rule set.Confusion matrix and vector image were used to judge the accuracy of extraction and superimposed.The results indicate that the rule sets put forward in this paper are suited to extract water conservancy facilities spatial element with GF-1 image.The extraction accuracy and superimposed accuracy of river are 97.1%and 94.1%,which are highest,of river blocking are 86.0%and 81.4%,which are lowest.Others have a higher accuracy.Thus,a good result can be obtained with the rule set of water conservancy facilities extraction with high resolution remote sensing image.The research results can be used to provide support to build sets of extracting water conservancy facilities and to update extraction of the water conservancy facilities based on GF-1 image.

关 键 词：GF-1影像 水利基础空间要素 特征提取 面向对象方法 规则集构建 

分 类 号：P237[天文地球—摄影测量与遥感]