Fusion of Multi-focus Images Based on the 2-Generation Curvelet Transform

Abstract

Curvelet is often used in image multi-scale analysis for its description ability to the image edges such as curve and line characteristics. On the basis of analyzing several common algorithms of image fusion, a new multi-focus image fusion method based on Curvelet transform is proposed according to the question of image fusion. The principle of Curvelet transform is described and the fusion rule of coefficients is analyzed. Firstly, two different focus images were decomposed using Curvelet transform separately, and then in the Curvelet domain of the two transformed images, the new Curvelet coefficients were acquired by adopting an efficient fusion rule, in which the low-frequency coefficients were integrated using the weighted average method, and high-frequency coefficients were integrated using regional energy analysis method. Finally, the fused coefficients are reconstructed to obtain fusion results. Experiments are conducted on multi-focus images by using different methods and the performances were evaluated with the indicators such as mean error, variance and information entropy. Experimental results show that the method is more suitable for multi-focus image fusion than some other ways, and the fusion image will have more information.