基于改进SURF算法的遥感图像配准

SURF算法用于遥感图像配准具有更好的鲁棒性和更高的效率。但由于遥感图像通常包含较多的重复特征,进行配准时容易产生误配,导致拼接图像出现重影和不精确。针对该问题,运用GTM方法对SURF算法进行改进后用于遥感图像配准。结合特征点的位置信息,分别构建特征点的图形,通过迭代使待配准图像的特征点图形一致,去除误配点。实验结果表明,使用改进后的算法进行遥感图像配准能有效去除误配点,实现准确匹配,从而实现正确拼接。

Remote sensing image registration based on improved SURF

SURF can achieve better robustness and efficiency when used to match remote sensing image.But remote sensing image often contain many repeated features,wrong matches are easily encountered when using SURF to match images,which result in incorrect mosaic.To resolve this problem,using GTM to improve SURF,combined with the location information of features,construct graphs based on features,attain accordance of graphs by iteration,and remove wrong matches.The results of experiments show that wrong matches are removed by using this advanced method.Correct matches are achieved,and resultantly accurate mosaic.