论文部分内容阅读
本文提出一种基于SIFT特征的不同源遥感影像自动匹配方法。首先利用Harris算子结合SIFT特征提取影像上均匀分布的特征点,建立高维SIFT特征描述符;然后以待匹配点与参考点间的欧氏距离为相似性测度,实施两种不同源遥感影像的特征匹配;最后将SIFT特征匹配结果作为初始值,采用将搜索范围扩大到尺度空间后的改进最小二乘影像匹配方法对SIFT特征匹配获得的同名像点进行精匹配。经对同一地区SPOT-5 HRG全色影像与航空摄影影像的匹配试验,取得了较好的结果。
In this paper, a novel SIFT-based automatic matching method for different remote sensing images is proposed. Firstly, the Harris operator and SIFT feature are used to extract feature points uniformly distributed in the image to establish a high-dimensional SIFT feature descriptor. Secondly, the Euclidean distance between the point to be matched and the reference point is a measure of similarity, and two different source remote sensing images Finally, the matching result of SIFT is taken as the initial value, and the matching of the SIFT feature matching is performed using the improved least squares image matching method which expands the search range to the scale space. Good results have been obtained by matching SPOT-5 HRG panchromatic images and aerial photography images in the same area.