An improved DCT based image fusion using saturation weighting and joint trilateral filter

Abstract

Image fusion is used to merge important information from several images into single image. Color artifacts and noise are the two major challenges for existing transform domain methods which reduce quality of the resulting image. This paper presents a novel method for multi-focus image fusion scheme based on ACMax DCT specially designed for visual sensor networks. The proposed technique consists of multi-focus image fusion technique based on higher valued alternating current coefficients computed in DCT in combination with saturation weighting based color constancy to reduce the color artifacts. The method of fusion may affect the edges and produce noise in the digital images so to overcome this problem joint trilateral filter has been integrated with proposed algorithm to improve the results. The experimental results verify that the proposed technique gives more correct explanation and has better quality when compared with other image fusion methods.