为了提高图像配准的质量,论文在分析常用图像配准方法的基础上,总结出根据特征进行图像配准的优势,进而使用 SIFT算法来提取特征点,使用高斯差分金字塔的分层构造确定和提取图像的空间极值点,从而得出精确而平稳的配准结果,为图像配准的优化与发展打下基础。%In order to improve the quality of image registration,based on the analysis of common image registration method,the advantage is summed up according to the characteristics of image registration.And the SIFT algorithm is used to extract the feature points,using Gauss differential to determine and extract the image’s spatial extreme points,thus the accurate and stable results are obtained to lay the foundation for the optimization and development of image registration.
展开▼