Research On The Algorithms Of Microscopic Image Fusion

The fusion of microscopic images is a reprehensive subject in image fusion domain. It is an effective method to extend the depth of focus and acquire a clear image of all relative focused objects. Consequently it could improve the utility ratio of image information and the reliability of object recognition. In this paper, we do the research of the fusion algorithms of microscopic images. To overcome shortcomings of performed image fusion algorithms in fusing images of micro-sample with high vertical extent, two new fusion algorithms are proposed.First, according to the characteristic of wavelet transform, a novel microscopic image fusion algorithm based on wavelet transform using Log-Gabor filter is proposed. This algorithm is based on wavelet transform, and Log-Gabor filter method based on region is used to design the fusion rule of approximate coefficients. This algorithm could prevent the blocking effect and avoid the blurry phenomenon of previous wavelet algorithm. Influences of different parameter settings on performance of fusion are investigated, and the result of experiment, which is compared to other well-known algorithms, indicated that this algorithm is effective and superior to the other algorithms.This paper also proposes a new microscopic image fusion algorithm based on BEMD. Because BEMD is a new form of fully two-dimensional multi-scale decomposition, it can decompose adaptively without selecting some parameters and has superior quality in extracting salient features in microscopic images. Here, PCA fusion rule and maximum selection or local significance fusion rule are applied to different image components obtained from the BEMD, so it can gain better fusion results. Experimental results demonstrate that the proposed algorithm is successful in multi-focus images fusion and significantly outperforms the traditional algorithms.