本文针对SIFT(尺度不变特征变换)算法存在的内存消耗多、运算速度慢的问题，采用金字塔和分块策略，首先对原始影像进行粗配准，然后进行分块影像匹配以实现精确配准。在匹配过程中，根据影像分辨率限制高斯金字塔影像的阶数，对特征点进行过滤；同时对匹配过程进行并行化，以提高算法效率。实验表明，改进算法在保证配准精度稳定的前提下，解决了原算法对内存要求高的问题，效率比原算法显著提高，适用于大范围遥感影像之间的配准。

According to the problems of large memory consumption and low computation speed in SIFT (Scale Invariant Feature Transform) algorithm, this paper adopts the strategy of pyramid and partitioning to register original images coarsely and matches partitioned images to realize accurate registration. During the process of matching, the number of Gaussian image octave is restrained according to image resolution and the feature points are fi ltered. Meanwhile, the matching process is paralleled to improve the effi ciency. Experiments show that on the premise that registration accuracy is stable, the proposed algorithm solves the problem of high requirement of memory and the effi ciency is improved signifi cantly, which is applicable for registering remote sensing images of large areas.