直接使用检测到的SIFT（Scale-Invariant Feature Transformation）特征点进行特征点匹配，匹配性能仍然有待提升.提出了改进的SIFT匹配算法，利用匹配特征点的尺度比直方图，估计出近似的图像尺度比k，然后将空间分辨率较高的图像初始尺度增大到k倍，再次提取特征点进行匹配.实验结果表明，相比于其它用尺度约束条件提升性能的匹配算法，基于初始尺度变化的SIFT匹配算法在处理结构型图像时性能得到了很大的提升.

Performance of the matching method which directly using SIFT feature points needs to be improved. This paper proposed an improved SIFT matching method which firstly uses the scale ratio histogram of the matched feature points to estimate the approximate scale ratio k of the image pair, then increases the base scale up to k times and extracts featur points points again for the higher resolution image, finally performs feature matching again to obtain the result. The experiment results show that, compared with the existing improved SIFT matching methods with scale difference, performance is improved significantly for structured image pairs.