摘要
遥感图像配准是遥感图像融合中最重要的预处理步骤。寻找适应性强、精度高、计算快的配准算法一直是研究的核心问题。在研究现有配准算法的基础上,提出了一种基于图像分类的特征匹配方法。该方法在基于控制点的多项式粗配准的基础上,利用分类图像相关实现了基于仿射变换模型的精配准。实验结果表明:对于同质和异质遥感图像,此方法的配准精度都达到了子像素级;而计算量与传统的基于灰度相关的方法和基于特征匹配的方法相比则大为减少。
The remote sensing image registration is one of the most important issues of image fusion. Researchers always focus on (searching) (adaptive,) accurate and fast registration algorithm. A feature matching (algorithm based) on classification is proposed, which uses the correlation of the classified images to realize a precise registration after a coarse (registration). (Experiment) (results) show the (feasibility) and (adaptability) of the method. To both (homogeneous) images and (heterogeneous) images, the (registration) (results reach) (sub-pixel) (precision). (The) (cost of) (computation) (is far) (less) than that of the (conventional) (methods) (through) (gray-based) (correlation) or (feature-based) (matching).
出处
《国防科技大学学报》
EI
CAS
CSCD
北大核心
2004年第2期35-40,共6页
Journal of National University of Defense Technology
基金
国家部委预研项目(413220202)
关键词
图像分类
粗配准
精配准
相关
image classification
coarse registration
precise registration
correlation
