Abstract: Image fusion technique has been frequently used to enhance the resolution of remote sensing data, while image fusion technique can cause the spectral distortion of resultant fusion images with the spatial detail features enhancing. Based on a review of current fusion algorithms for remote sensing image, this paper proposes an improved algorithm by using FIHS（Fast Intensity-Hue-Saturation） and BT（Brovey Transformation）, it’s named the FIHS-BT algorithm. The multispectral image and low-pass filtered panchromatic image are calculated to simulate information of panchromatic image. This can help remove the low frequency information of panchromatic image and the high frequency information of multispectral image. To reduce the degree of saturation is stretched or compressed, the fusion image is generated by the square root of the product of FIHS and BT. The FIHS-BT can fuse an image without reference band and the limitation of band number, and the FIHS-BT can keep the best balance between spectral fidelity and the ability of gaining high frequency information. A subset from the IKONOS image is used as test data. The several fusion algorithms have been used to fuse the test image, such as IHS, BT, SVR, and so on. The results show that the spectral fidelity and high frequency information of the FIHS-BT algorithm are generally better than the algorithms of FIHS and BT.

Key words: image fusion, spectral fidelity, high frequency information, FIHS-BT

摘要： 影像融合可以显著提高影像的空间分辨率，但融合时高频信息的注入势必会造成融合影像的光谱失真。在对FIHS（Fast Intensity-Hue-Saturation）算法和BT（Brovey Transformation）算法光谱失真原理研究的基础上，提出一种能有效减弱光谱失真和提高影像空间分辨率的FIHS-BT算法。该算法通过利用全色影像的低频信息和多光谱影像生成模拟全色影像，减弱全色影像低频信息对融合影像光谱信息的影响，削弱多光谱影像高频信息对融合影像高频信息的干扰；然后，采用FIHS和BT乘积的平方根生成融合影像，以减小饱和度在变换中被拉伸或压缩的程度。选取IKONOS影像为数据源，采用BT等九种融合算法与FIHS-BT算法进行融合比较实验，并对融合结果从光谱保真度和高频信息融入度两个方面进行定性和定量评价。实验结果表明，FIHS-BT算法在光谱保真度和高频信息融入度方面较FIHS算法和BT算法均有显著改善。

关键词: 影像融合, 光谱保真度, 高频信息, FIHS-BT