A Comparative Study on the Quality of Different Image Fusion Methods Based on SPOT4 and Landsat5 TM
[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.
NETL
NSTL
万方数据
