A New Wavelet-based Multifocus Image Fusion Technique using Method Noise-Median Filtering

摘要

Multifocus Image fusion is a technique to combine two or more different multi-focused images to generate an all-in-one focus image. In this paper, a new wavelet-based multi focus hybrid technique using method noise is proposed. Firstly SWT is applied to both the images to decompose the images in respective four subbands (LL, LH, HL, and HH). LL (low) is the approximation coefficient and others are LH, HL and HH are detailed coefficients for the source image. Then after that with the help of Inverse SWT we reconstruct the enhanced decomposed coefficients. When we compare the fused image with source image we get some residual part known as method noise. After processing i.e. filtering the method noise part, the enhance fused image is produced which is better than previous without method noise and other previous mentioned available literature. The proposed hybrid approach is used to generate a better enhanced and less distorted image that contains the full information of both degraded input images. The experimental results show that the proposed approach is better than the previous literature related to wavelet in terms of qualitative and quantitative results.