An image fusion algorithm based on shearlet

Abstract

We put forward a new method for multi-focus image fusion based on the shearlet transform. It is combined with the multi-scale method for capturing the geometry features of the multidimensional data. Firstly, the shearlet transform is adopted to classify the original image in both the scales and directions to get the low and high frequency sub-band coefficients. Secondly, based on the coefficients, we analyze the relationship between the low frequency coefficients and the sharpness of the image, a weighted averaging scheme is presented. Finally, the high frequency coefficients are obtained by selecting the corresponding larger absolute value. Experimental results show that the proposed method is feasible with the better quality than the state-of-the-art one.