中文说明：SIFT是一种基于特征的配准方法。SIFT特征匹配算法是 DavidG.Lowe在2004年总结了现有的基于不变量技术的特征检测方法的基础上，提出的一种基于尺度空间的、对图像缩放、旋转甚至仿射变换保持不变性的特征匹配算法。该算法匹配能力较强，能提取稳定的特征，可以处理两幅图像之间发生平移、旋转、仿射变换、视角变换、光照变换情况下的匹配问题，甚至在某种程度上对任意角度拍摄的图像也具备较为稳定的特征匹配能力，从而可以实现差异较大的两幅图像之间的特征的匹配.该程序通过提取两图特征值，进行特征值匹配。
English Description：
SIFT is a feature-based registration method. SIFT feature matching algorithm is summarized DavidG.Lowe invariant feature detection technique based on the basis of the existing 2004, a proposed space-based scale, image scaling, rotation and even keep invariant affine transformation feature matching algorithm. The algorithm matching ability, stable characteristics can be extracted, can handle translation between two images, rotation, affine transformation, perspective transformation, lighting transform matching case, even to some extent arbitrary angle shooting the image also has a more stable feature matching capability, which can achieve a large difference in matching features between two images.The program features two plans by extracting value, the feature values match