论文部分内容阅读
本文针对大幅影像配准时计算量大、配准点分布严重不均匀以及错配率高等问题进行了研究,提出一种可保持配准点均匀分布的快速配准算法。首先对基准影像和待配准影像进行分块提取SIFT特征算子,然后局部自适应地对影像块中的匹配点进行几何约束,从局部选取最优匹配点以达到提高图像匹配准确率的目标。通过对大量不同类型的影像进行配准实验,实验结果表明,当图像存在传感器差异甚至较高分辨率差异时,本文算法也能够快速准确地达到配准目的。
In this paper, aiming at the problem of large amount of image registration, large distribution of registration points and high mismatch rate, a fast registration algorithm which can keep the registration points evenly distributed is proposed. Firstly, the SIFT feature operator is extracted by segmenting the reference image and the image to be registered, and then the matching points in the image block are locally and locally constrained geometrically, and the optimal matching point is selected locally to achieve the goal of improving the image matching accuracy . By registering a large number of different types of images, the experimental results show that the proposed algorithm can achieve the target of registration quickly and accurately when there are differences in images and even higher resolution of the images.