节点文献
基于SIFT和渐进局部几何约束的多源遥感图像自动匹配研究
Automatic matching for multi-sources remote sensing imagery based on SIFT and heuristic local geometric constrained strategy
【Author】 AN Ru~1 WANG Hui-lin~2 YAN Xiang~1 1.Geo-informatics Department of College of Hydrology and Water Resources.Hohai University,Jiangsu Nanjing 210098; 2.College of Geography and Sea Sciences,Nanjing University,Jiangsu Nanjing 210093
【机构】 河海大学水文水资源学院地理信息科学系； 南京大学地理与海洋科学学院；
【摘要】 本文研究了通过自动寻找同名特征点进行多源遥感图像匹配的方法。运用 SIFT 进行特征点提取与描述,获得初始匹配的特征点对;对于初始的错误匹配点对,采用渐进的几何约束策略进行剔除,进而得到用于匹配纠正的同名特征点对。应用多源遥感图像进行了匹配实验,表明该方法具有尺度、旋转和光照不变的特点,是一种有效的匹配方法。
【Abstract】 A method for automatic matching for multi-sources remote sensing imagery based on SIFT(Scale Invariant Feature Transform)and heuristic local geometric constrained strategy is investigated.SIFT is used to key-points detection and description,and initial matched key-point pairs is obtained.A heuristic local geometrically constrained strategy is employed to remove mis-matched key-point pairs in initial matching stage,and homonymous key-points pairs are derived for next image matching.Multi-sources remote sensing imageries are used to experiment,and it is manifest that the approach in this paper is an efficient matching method with invariant characteristics of scale,rotation and illumination.
【Key words】 Image matching； SIFT； feature point detection； feature point description； geometric constraint；
- 【会议录名称】 第十四届全国图象图形学学术会议论文集
- 【会议名称】第十四届全国图象图形学学术会议
- 【会议时间】2008-05
- 【会议地点】中国福建福州
- 【分类号】TP391.41
- 【主办单位】中国图象图形学学会