基于角点检测法的图像配准

采用角点检测算法,对图像中任一像素点和以它为中心的圆形区域USAN进行图像配准,将边缘提取和轮廓跟踪过程融为一体,并通过沿圆弧曲线扫描获取角点信息建立图像间角点的对应关系,由此得到初配准参教,且使用迭代过程以提高配准的精度.试验证明,该算法对图像间的旋转角度没有限制,配准精度高而且计算量小.

Image Registration Based on Corner Detection Method

Based on corner detection algorithm, image registration is for any pixel of an image and centering on its circular area USAN. The process makes the image border pick-up with its outline tracking an integral whole. Corner information among images can be achieved by scanning along the circle. After the corners of different frames of the same scene are detected, the matching corner pairs between these frames can be found. Based on these pairs, initial image registration parameters can be estimated. Use iterative process to improve image registration accuracy. Experiment proved that, this algorithm has no limitation on rotation degree between images, and image registration precision is high while the computation load is small.