A Statistical Analysis of Multiresolution Fused Images to Enhance the Image Quality

Abstract

Image enhancement is an important tool in digital image processing, which is primarily used for better image visual quality. Image fusion attains attention because of its ability to produce enhanced image as resultant output by fusing multiple spectrums information (i.e. red, blue, Near Infra Red (NIR) and green) in reliable way. Line features visibility is excellent in blue and red bands while vein features are revealed in red band. Both vein structural information and partial line information are shown by near infra red (NIR) respectively. Multi-spectral imaging is proposed in proposed work to acquire accurate information compare to traditional methods. Wavelet transform is regarded as successful statistical transform technique in image fusion process. But it frequently suffers from discontinuities for limited number of coefficients. To overcome the discontinuities issue in effective way by  curvelet is proposed in place of wavelet for better fusion efficiency.The simulation results are compared with different performance measurements for evaluation and the proposed curvelet method yields better results than popular wavelet statistical methods.