论文部分内容阅读
针对多源遥感影像间几何变形和灰度差异造成的匹配困难问题,提出一种结合SIFT和边缘信息的影像匹配方法。首先在高斯差分尺度空间进行特征点检测,并采用相位一致性提取可靠的边缘信息;然后结合改进的SIFT和形状上下文对特征点进行描述;最后将欧氏距离和χ2统计作为相似性测度获取同名点。相比于SIFT算法,本文方法可有效地提高匹配正确率,并获得更多的同名点。
Aiming at the difficulty of matching caused by geometric distortion and gray level differences between multi-source remote sensing images, an image matching method based on SIFT and edge information is proposed. Firstly, the feature points are detected in Gaussian differential scale space, and the reliable edge information is extracted by phase consistency. Secondly, the feature points are described based on the improved SIFT and shape context. Finally, the Euclidean distance and χ2 statistics are taken as the similarity measures point. Compared with the SIFT algorithm, this method can effectively improve the matching accuracy, and get more points with the same name.