Fusion of Multisensor Data: Review and Comparative Analysis


This paper reviews and analyses seven fusion techniques: IHS, BT, HPF, HPM, PCA, FT and CA. The performance of each method is determined by two factors: how the low resolution PAN image is computed and how the modulation coefficients are defined. If the low resolution PAN image is approximated from the low resolution MS image, it usually has a weak correlation with the high resolution PAN image, leading to color distortion in the fused image. If the low resolution PAN is a low-pass filtered high resolution PAN image, it usually shows less spectral distortion. If the modulation coefficient is set as a constant value, the reflectance differences between the PAN and the MS bands are unaccounted, and the fused images bias the color of the pixel toward the gray. By combination of the visual inspection results and the quantitative results, it is possible to see that the experimental results are in conformity with the theoretical analysis. HPF method followed by HPM produces the synthesised images closest to those the corresponding multi-sensors would observe at the high-resolution level.