Multi-focus image fusion based on differential evolution algorithm has been demonstrated to be a simple and efficient method. However, several some problems remain. First, the fusion rulesin which the best block size is selected is only based on the original differential evolution algorithm, which easily falls into the local convergence and thus affects the global search ability. When the sharpness values of the corresponding blocks are equal, the method becomes ineffective, because the block effect is enhanced. Secondthe algorithm ignores image size. Thus, calculating larger pictures becomes complex and time consuming. Clear and fuzzy areas are further divided, thus causing unnecessary calculations. Therefore, a multi-focus images fusion method based on an improved differential evolution algorithm and adaptive block mechanism is presented. First, the source images are divided by a fixed size once. Then, the boundary region is searched to find adaptive blocks using the improved differential evolution algorithm. If the sharpness values of the corresponding blocks are equal, the extends block mechanism is applied to determine the block with the highest sharpness value. Experimental results show that the improved algorithm can obtain better fusion effects and consume less time compared with the original differential evolution algorithm.
Digital Object Identifier (DOI)
Yong, Feng; Tiezhu, Li; and Shangbo, Zhou
"Multi-focus Image Fusion using an Improved Differential Evolution Algorithm and Adaptive Block Mechanism,"
Applied Mathematics & Information Sciences: Vol. 08:
5, Article 35.
Available at: https://digitalcommons.aaru.edu.jo/amis/vol08/iss5/35