Paper Title
An Enhanced Framework For Sharpenig The Multispectral Images Using Genetic Algorithm

Abstract
There are a number of applications in satellite remote sensing that require images with high spatial and reasonable spectral information. The fusion of the multispectral (MS) and panchromatic (pan) images or pan sharpening provides a high resolution (HR) colored image by merging the clear geometric features of the HR pan image and the color information of the MS bands. This HR pan image can be created by applying reconstruction techniques on the high-pass filtered bands of the MS image by utilizing the sub-pixel shifting between bands. In the proposed framework, the Genetic Algorithm (GA) is used for optimal estimation of the sub-pixel shifts between bands, blurring parameters for image reconstruction and adaptive weights for image fusion. The HR pan image is generated by GA-based Projections onto convex sets method. GA is considered global search methods. Therefore, it differs from those methods based on conventional high computation iterative solution of equations.Experimental results demonstrates an accurate sub-pixel image registration and animproved spatial quality of MS image by optimal estimation of the blur and the fusion weights for keeping a reasonable color appearance. Appropriate constraints improve the convergence rate of the GA-based registrationmorethan the classic methods. Indexterms- Satellite image, Registration, Restoration, HPF Fusion, Sharpening, GA.