目前多种成熟的融合算法已经应用在各种遥感软件中，但是融合方法的选择往往会因融合对象的不同而有 所差异。为了评价出各个融合算法在QuickBird影像融合上的优缺点，本文在像素级的融合层次上运用多尺度分 析的方法进行了融合实验。实验中，根据不同的算法原理引入了七种常用的融合算法，并以空间细节信息、光谱质量以及亮度 信息作为统计参数，对实验数据进行了比较研究，分析出了几种融合方法的差异。研究表明，一些传统的融合方法如PCA变换和IHS变 换已不适用于QuickBird这种高分辨率影像的融合，而基于小波的PCA变换、小波变换以及HPF变换在实验中有较好的表现。本次 实验也为其它高分辨率卫星遥感影像的融合工作提供了参考。

Although more mature fusion algorithms have been used in various kinds of remote sensing softwares, the choice of fusion algorithms is often different for different fusion objects. To evaluate the advantages and disadvantages of each algorithm used in the fusion of Quickbird images, the image fusion experiment is made at the image pixel level by using a multi-scale method. In the experiment, seven common fusion algorithms are incorporated according to different principles. Their experimental data are compared with each other and the differences of several fusion methods are analyzed. The research shows that some traditional fusion methods, such as PCA transform and HIS transform, are no longer suitable for the fusion of Quickbird high resolution images. Instead, the wavelet-based PCA transform, wavelet transform and HPF transform methods have good behaviors in the experiment. This experiment is also helpful to the fusion of the high resolution images obtained by other remote sensing satellites.