基于特征点的微波遥感图像自动配准算法

Microwave Remote Sensing Image Automated Registration Algorithm Based on Feature Point

摘要: 图像配准是图像处理工作（图像融合、图像镶嵌等）的核心技术。本文提出了一种有效的遥感图像自动配准算法，该算法采用改进的Harris 算子提取角点特征，利用待配准图像之间的灰度相关性进行粗匹配，然后运用虚拟三角形对全等的准则进行精匹配得到最终的匹配角点，最后通过这些角点求得刚体变换模型参数。实验结果表明：该算法是正确和有效的。

Abstract: Image registration is crucial technologies of image processing. In this paper, a reliable method of automatic remote sensing image registration is proposed, an improved technique based on Harris is used to extract corners ,and the correlation of the gray level of the images is used to rough matching；Secondly ,obtain all matched corners according to the congruent virtual triangles ;Finally ,estimate the optimal parameters of rigid transformation model by them. The experiment results show that the proposed method is accurate and effective.