Application Research of Fused Image of CBERS-02 and SPOT5 Data in Land Use Monitoring of Coastal Zone
CBERS-02与SPOT5融合数据及其在海岸带土地利用调查中应用能力综合评价

In this article,taking Hengqin island as experiment region,Normal image fusion algorithms (HIS transform,Brovey transform,PCA transform,wavelet algorithm,SFIM transform and PANSHARP transform) are used to fuse the multispectral image of CBERS-02 and panchromatic image of SPOT5.In order to analyze each band of fused images and find the best fusion method by criterion,four quantitative indexes such as the average value,the entropy,the correlation coefficient,the standard difference and the gradient are computed and analyzed.Based on the PANSHARP-fused image,the land use categories are extracted by support vector machine classifier,the total classify precision is 82.55% and Kappa index value is 0.7358.The results show that by fusing the multispectral image of CBERS-02 and panchromatic image of SPOT5,both spatial and spectral resolutions can be improved,and the fused images are of wide application in various aspects such as land use monitoring and evaluation.