一种基于兴趣点方向特征的图像拼接方法

提出了一种基于兴趣点方向特征的图像拼接算法IPOF(Interest Point Orientation Feature),该算法利用Harris角检测器提取出两幅图像的兴趣点并为每个兴趣点分配一个主方向,采用方向相关系数法提取出初始匹配对,根据特征点之间的关系去除伪匹配对,得到两幅图像的对应兴趣点特征对从而确定变换参数,最后使用加权平均的方法融合图像。实验表明,该算法在图像间存在任意角度的旋转及平移的情形下,能有效地实现图像的平滑镶嵌。

Image mosaic method based on orientation feature of interest point

This paper proposes a novel algorithm for image mosaic based on interest points.Interest points are firstly extracted by Harris comer detector in two images and each interest point is assigned one dominant orientation.Using orient correlative coefficient matching method to extract the initial matches,then reject the false match according to the relationship among feature points.Transform parameters are determined by the matches,and weighted average method is used to blend images.The experimental result indicates that the proposed algorithm can effectively stitch two images which are different by a rotation and a translation.