基于SPOT4和Landsat5 TM的不同影像融合方法质量对比研究

A Comparative Study on the Quality of Different Image Fusion Methods Based on SPOT4 and Landsat5 TM

【目的】探究不同影像融合方法对SPOT4全色影像和Landsat5 TM多光谱影像融合效果的影响。【方法】采用13种不同的影像融合算法开展融合试验,并通过定性和定量分析的方式,对融合结果进行综合评价分析。【结果】结果表明,融合影像的信息熵相比原影像有所增强,RGB三通道的熵值分布更为均匀,主要集中在50~200之间。此外,不同地物类型对融合方法的效果产生了显著影响。对于水体和植被,Brovey、HSV、IHS和EF方法展现出了良好的光谱保真性和细节增强能力,因此更适用于目视解译。NND法在这些区域出现了轻微的光谱失真现象,导致其分类精度偏低。【结论】针对不同传感器的影像选择适当的融合方法,对于提高遥感影像的解译精度和应用价值具有重要意义。

[Purposes]This paper aims to explore the influence of different image fusion methods on the fusion effect of SPOT4 panchromatic image and Landsat5 TM multispectral image.[Methods]Twelve dif-ferent image fusion algorithms were used to carry out fusion experiments,and the fusion results were comprehensively evaluated and analyzed by visual interpretation and quantitative evaluation.[Findings]The research results show that the information entropy of the fused image is enhanced compared with the original image,and the entropy distribution of RGB three channels is more uniform,mainly concentrated in the range of 50~200.In addition,different types of features have a significant impact on the effect of the fusion method.For water and green vegetation areas,Brovey,HSV,IHS and EF methods show good spectral fidelity and detail enhancement ability,so they are more suitable for visual interpretation.How-ever,the NND method has slight spectral distortion in these regions,which leads to its low classification accuracy.[Conclusions]To sum up,the selection of appropriate fusion methods for images from different sensors is of great significance to improve the interpretation accuracy and application value of remote sensing images.