Fusion of Multisensor Data:Review and Comparative Analysis

Image fusion is a formal framework which is expressed as means and tools for the alliance of multisensor, multitemporal, and multiresolution data. Multisource data vary in spectral, spatial and temporal resolutions necessitating advanced analytical or numerical techniques for enhanced interpretation capabilities. This paper reviews seven pixel based image fusion techniques - intensity-hue-saturation, brovey, high pass filter (HPF), high pass modulation (HPM), principal component analysis, fourier transform and correspondence analysis.Validation of these techniques on IKONOS data (Panchromatic band at I m spatial resolution and Multispectral 4 bands at 4 in spatial resolution) reveal that HPF and HPM methods synthesises the images closest to those the corresponding multisensors would observe at the high resolution level.