国土资源遥感Issue(3)：48-54,7.DOI:10.6046/gtzyyg.2014.03.08
基于特征点与边界信息的全自动多模态遥感图像配准方法
Automatic approach based on point and curve features for multimodal remote sensing image registration
摘要
Abstract
Due to the significant spectral difference in multimodal image registration, the rate of correctly matched feature- points calculated with point - feature detecting and matching algorithms, such as speed up Robust features,SURF, is quite low in some difficult cases. In order to solve this issue, this paper proposes a novel image registration approach with the help of point and curve features. It adopts the location information indexed by the correct matching pair of feature-points and transformation information determined by the pair of correct matching curves simultaneously. Also, the feasibility and advantages of this algorithm were all confirmed by the experiments in this paper. The results show that this method can achieve automatically the registration of the multi -modal remote sensing images in the completely unattended case and align some kinds of remote sensing images automatically. In addition, it is more robust and reliable.关键词
多模态遥感图像/图像配准/特征点检测与匹配算法/曲线匹配算法Key words
multimodal remote sensing images/image registration/point -feature detecting and matching algo-rithm/curve-feature matching algorithm分类
信息技术与安全科学引用本文复制引用
宋智礼..基于特征点与边界信息的全自动多模态遥感图像配准方法[J].国土资源遥感,2014,(3):48-54,7.基金项目
上海市自然科学基金项目“遥感与地理信息系统中全自动高精度图像配准新技术的研究冶(编号:12ZR1431000)、上海应用技术学院科学技术发展基金项目“医学、遥感图像配准技术及其应用的研究冶(编号：YJ2011-68)和085项目“机器嗅觉冶(编号:405ZK124127)共同资助。 （编号:12ZR1431000）
