Image Fusion Based on Spatial Weightage in Nonsubsampled Contourlet Transform Domain

Abstract: Satellite imageries are categorized as both panchromatic and multispectral images. The panchromatic image has higher spatial resolution while the multispectral image has relatively lower spatial resolution but are rich in spectral information .The fusion of these two can produce a better imagery . Most of the fusion methods available to merge a high- resolution panchromatic image and low-resolution multi-spectral images may distort the spectral characteristics of the multi-spectral image or the spatial resolution of he panchromatic image may reduce. In this paper, a fusion algorithm for multi-spectral and panchromatic images based on spatial weightage called region correlation coefficient and in the Nonsubsampled Contourlet Transform (NSCT) Domain is proposed. The proposed fusion idea partitions both the source images into regions and Region Correlation Coefficient (RCC) is calculated between the corresponding regions of multispectral and panchromatic images .This RCC determines the regions of multispectral image with better spatial resolution and those with better spectral quality. The fusion is performed in NSCT domain. For that NSCT is performed on individual regions of panchromatic image and intensity component of multispectral image at different scales and directions. The obtained NSCT coefficients are fused taking into consideration the spatial weightage factor. Performance analysis show that the algorithm proposed is better than the IHS transform method and simple NSCT method.

Keywords: Satellite imagery, multispectral image, panchromatic image, Nonsubsampled Contourlet Transform, IHS transforms, Region Correlation Coefficient