改进支持向量机的遥感影像道路提取技术研究  

Research on Road Extraction from Remote Sensing Image Based on Improved Support Vector Machine

作　　者：王艳梅[1] 李金雨 冯海霞 WANG Yan-mei;LI Jin-yu;FENG Hai-xia(Traffic Engineering College,Anhui Sanlian University,Hefei 230601,China;School of Transportation and Logistics Engineering,Shandong Jiaotong University,Jinan 250357,China)

机构地区：[1]安徽三联学院交通工程学院,安徽合肥230601 [2]山东交通学院交通与物流工程学院,山东济南250357

出　　处：《浙江水利水电学院学报》2021年第3期74-76,86,共4页Journal of Zhejiang University of Water Resources and Electric Power

基　　金：安徽省自然科学基金资助项目(KJ2019A0902)。

摘　　要：道路信息是重要的地理信息,是城市地理信息数据库的重要组成部分。随着高空间遥感影像的出现,如何从遥感影像中准确地提取道路信息成为一个重要课题。据此,通过研究SVM道路提取方法,针对SVM提取道路信息时存在的问题,提出了K均值聚类与SVM相结合的道路提取方法。通过实验结果显示,K均值聚类与SVM相结合的道路提取方法比直接使用SVM提取道路的方法准确度更高,冗余误差和遗漏误差更低。Road information is an important geographic information and a vital part in urban geographic information database.Due to the development of high-space remote sensing images,how to accurately extract road information from remote sensing images has become an important topic.By studying the SVM road extraction method and its problems existing in road information extracting,a road extraction method combining K-means clustering and SVM is proposed.The experimental results show that the road extraction method combining K-means clustering and SVM has higher accuracy,lower redundancy error and omission error compared with SVM only.

关 键 词：K均值 SVM 道路提取 遥感影像 

分 类 号：P237[天文地球—摄影测量与遥感]