摘要： 提出了一种新的基于独立成分分析的全色图像和多光谱图像的融合算法。对于两幅图像的重叠区域，使用独立成分分析去除遥感影像中高阶数据冗余问题，然后对通过独立成分分析所分离的各个分量进行叠加从而得到最终的融合结果。该方法的优点在于去除了原始图像上的数据冗余，提高了融合后图像的信息量和信噪比。

关键词: 多光谱融合；独立成分分析；全色图像

Abstract: In this paper, a new approach using the independent component analysis (ICA) for remote image fusion between panchromatic and multi-spectral images is presented. First, it performs ICA to eliminate high-order image data redundancy for two overlapping images from different channels, then the respective independent components are fused. The advantage of this method lies in the reduction of data redundancy, and image quality and signal-to-noise ratio are improved.

Key words: Multi-spectral image fusion; Independent component analysis; Panchromatic image