摘要
在模版匹配的基础上,提出一种基于二次曲面拟合的配准方法.在配准点的邻域内把序列图像的相似性测度拟合成二次曲面函数,通过求二次曲面函数的极大值点(最佳匹配点),得到配准图像的亚像素位移,并用该位移来估计遗失帧,然后用插值法得到高分辨率图像.试验结果验证了该方法的有效性.
On basis of pattern plate registration,a method to fit similarity measure of the sequence image to quadratic surface function in neighborhood of the registration point was presented. Through resolving maximum value of the quadratic surface function （the optimum registration point）, sub-pixel accurate shift of the registering images could be obtained. By using the sub-pixel shift information to estimate missing frame,the high resolution images could be gained from interpolation. Feasibility of the new methods was proved by the experiment.
出处
《内蒙古师范大学学报（自然科学汉文版）》
CAS
2007年第4期446-449,共4页
Journal of Inner Mongolia Normal University(Natural Science Edition)
基金
广东省自然科学基金资助项目(06301315)
关键词
高分辨率
图像配准
图像复原
亚像素
二次曲面拟合
high-resolution
image registration
image reconstruction
sub-pixel
qadratic surface fitting
