一种基于遗传算法的自适应多聚焦图像融合新方法

An adaptive multi-focus image fusion method based on genetic algorithm

提出一种基于遗传算法的自适应分块多聚焦图像融合新方法,该融合算法的主要思想是选择拉普拉斯能量和sum-modified-Laplacian(SML)作为衡量图像子块清晰的度量,边缘信息保留值作为适应度函数,最终运用遗传算法自动搜寻到最优子块大小实现图像的融合。实验结果表明:本文提出的融合方法性能明显优于传统的图像融合方法。

A novel adaptive block-based multi-focus image fusion method was presented by using genetic algorithm.In the method,the clarity of image sub-block was measured by the sum-modified-Laplacian evaluation metric,and calculated the fitness function of each individual through edge information retention.Then via genetic algorithm,the optimal size of the sub-block was automatically obtained,which was finally used to fuse the source images.Experimental results indicate that the performance of our method is obviously superior to that of series of conventional fusion methods.