随着CBERS-02B星(以下简称02B星)多光谱CCD数据与全色HR数据在各个领域的广泛应用,两者相差较大的空间分辨率给图像融合带来难题.在对改进型的IHS变换、主成分变换、Brovey变换、GS变换等融合算法分析的基础上,利用02B星数据进行融合试验,并对试验结果进行定性与定量评价,结果表明:①Brovey变换融合效果较差；②在影像的阴影区域,改进型的IHS变换融合结果存在光谱失真；③PCA变换光谱保真较好；④GS变换无论是统计结果还是图像的色彩和清晰度都明显优于上述三种方法.%With the wide application of CBERS-02B Satellite (named as 02B Satellite as follows) multi-spectral CCD data and HR panchromatic data in various fields,it brings the difficulty in image fusion due to the large difference of spatial resolution.Based on the analysis fusion algorithms such as improved-IHS transformation,principal component transformation (PCA),Brovey transformation,GS transformation,etc,02B satellite data is used to for fusion test and the test result is evaluated by the qualitative and quantitative indexes.The results show that Brovey fusion is a less effective transformation method,improved-IHS transformation has the spectrum distortion in the shade area of image and the relationship between the different spectral changes significantly,PCA transformation spectral fidelity is better,and GS transformation is significantly better than the above three methods in terms of the statistics or the color and clarity of the image,but the image brightness need to be improved.
展开▼
