一种自适应极坐标变换的旋转图像配准方法

Rotated image registration using adaptive polar transform

对数极坐标变换具有较好的旋转不变性,但是同时存在采样不均匀的问题:在靠近中心处的过采样和靠近边缘处的欠采样.针对这一问题,提出一种自适应极坐标变换的方法,从笛卡尔坐标系转换到极坐标系时,实现图像的均匀采样,其采样点数目与对数极坐标变换相比,减少近50%,计算量大幅度减少.为了保持对数极坐标变换的旋转不变性,采用投影变换方法将极坐标变换后的图像从二维降到一维.实验表明该算法实时性好、精度高.

Log-polar transform（LPT） is a well known tool for image processing for its rotation invariant properties.However,it suffers from nonuniform sampling e.g.oversampling at the fovea area and undersampling at the peripheral area.Aiming at these questions,the author introduces adaptive polar transform（APT） technique in the spatial domain.Applying the projection transform the transformed image is reduced from 2-D to 1-D vector.The proposed method yields real time and more accurate registration than that of the LPT.