基于SIFT和NCC的多源遥感影像配准方法

针对多源遥感影像的配准，提出了一种结合SIFT算法和归一化互相关（NCC）匹配算法的配准方法。该方法采用SIFT算法提取特征点并进行匹配得到一定数量的特征点对后，利用SIFT特征点的尺度和方向信息对NCC进行改进，进一步从未能匹配的特征点中获取匹配点对，经粗差滤除后得到有效的匹配特征点对，随之进行影像配准。方法结合了SIFT算法和NCC算法的优点，解决了多源遥感影像因辐射差异和几何差异造成的难以正确配准的问题。实验结果表明，算法具有较强的鲁棒性，并取得了较好的配准精度。

Multi-sensor image registration method based on SIFT and NCC algorithm

To resolve multi-source remote sensing image registration,a new method combining Scale Invariant Feature Transform （SIFT） algorithm and Normalized Cross Correlation （NCC） is proposed.This method adopts SIFT algorithm to match feature point and to get a certain amount of feather points,then uses the scale of SIFT feature points and directional information to improve NCC in order to get the matching point.After gross filtering the effective matching feather points,the image registration can be done.This algorithm combines the advantages of SIFT and NCC,and solves the problem that the multi-source remote images are difficult for correct registration because of geometric differences and radiometric differences.Experimental results show that this algorithm has strong robustness,and achieves high registration accuracy.