Fusion Techniques For Multispectral And Panchromatic Images Analysis

Fusion Techniques for Multispectral and Panchromatic Images
Abstract:
Image fusion is the process of merging two or more images obtained from the same sensor at different times or from two or more sensors at the same instant. The objective is to obtain more information from the fused image than from the individual images. In satellite images, the lower spatial resolution multispectral images are fused with higher spatial resolution panchromatic images. The fusion should result in the transfer of spectral and spatial information without introducing any artifacts. The goal is to combine the spectral characteristic of the low-resolution data with the spatial resolution of the panchromatic image and obtain a high-resolution multispectral image.
The most commonly recorded image types in a satellite are typically multispectral images panchromatic images. While the panchromatic images record the total intensity of the radiation falling on each pixel, the multispectral images record the intensity of radiation in small band of visible spectra, including the RGB and the infrared region [4]. High spatial resolution is provided by panchromatic images and high spectral resolution is provided by multispectral images. While the multispectral image can be considered as having low spatial resolution but accurate color data the panchromatic image can be considered as having higher spatial resolution but a gray scale image. The applications of remote sensing need both high spectral and high spatial information for providing more information on the area being imaged. The fusion of multispectral and panchromatic images is thus done to provide more information on the scene than the individual MS and PAN images separately. Image fusion can in general be done at different levels namely pixel level, feature level, object level and decision levels. The level of fusion to be used depends on the intended application of the fused image [1], [2]. Many fusion techniques have been proposed for the fusion of multispectral and panchromatic images. The most widely used fusion techniques were intensity-hue-saturation (IHS), high pass filtering,
The disadvantage observed in this method is the color distortion when compared with the original multispectral image. Several modifications have been proposed to overcome this difficulty. One such method proposed by Tu et al in 2005 proposes to use four bands with the fourth band being an infrared component. This resulted in lesser color distortion since it allowed the calculated intensity of the multispectral image to better match the panchromatic image