基于区域能量的多聚焦图像融合算法  

Multi-focus image fusion algorithm based on regional energy

作　　者：王建[1] 刘肖[1] 王国珲[2] 

机构地区：[1]西安航空学院计算机工程系,陕西西安710077 [2]西安工业大学光电工程学院,陕西西安710032

出　　处：《陕西理工学院学报（自然科学版）》2012年第5期39-43,共5页Journal of Shananxi University of Technology:Natural Science Edition

基　　金：陕西省教育厅科研计划资助项目(11JK0996)

摘　　要：针对同一场景两幅严格配准的多聚焦图像的融合问题,设计了一种基于区域能量的多聚集图像融合算法。该算法利用金字塔变换得到图像的多分辨序列,采用基于区域能量的融合方法,在图像的相应层次序列的各级金字塔图像上进行融合,以获取最终的融合图像。从目视效果来看,采用本文算法进行图像融合,能够将两幅多聚焦图像融合成一幅比较清晰的单一聚焦图像。实验结果表明,本文算法与同类其它算法相比较,熵提高了0.01%～4.4%,交叉熵降低了52.93%～95.77%,互信息提高了3.13%～26.55%。Focusing on the fusion problem of the multi-focus images from same scene,a novel multi-focus image fusion algorithm based on regional energy was designed.By using multiresolution image sequences obtained from the pyramid transform,the source images were fused on each corresponding levels with the regional energy method,then the fused image was achieved through inverse pyramid transform.From the visual effect,the proposed algorithm can fuse into a clear single focus image from two-focus image.Experimental results and comparisons with three traditional algorithms demonstrate that the proposed algorithm is superior with 0.01% to 4.4% improvement of entropy,52.93% to 95.77% reduction in cross entropy and 3.13% to 26.55% enhancement in mutual information.

关 键 词：多聚焦图像 金字塔变换 图像融合 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]