摘要： 为了获得同一场景内所有物体都清晰的图像，提出一种新的多聚焦图像融合算法。把待融合图像进行分块，构造融合块的清晰度评价函数(区域锐度)，根据其区域锐度值，判断融合块应取自哪幅源图像。采用投票选举的方法对融合图像进行一致性校验，对相邻但来自不同聚焦图像的融合块进行加权融合。实验结果证明，与基于小波分解的融合算法相比，该算法速度快、效果好。

关键词: 图像融合, 多聚焦图像, 清晰度, 区域锐度

Abstract: In order to get a clear image that contains all relevant objects in an area, this paper proposes a new multi-focus image fusion algorithm. It divides the candidate images into small blocks, and constructs the blocks region acutance function to evaluate image definition. What source images the fused block obtains from depends on the region acutance value of the corresponding block. In the endgame, consistency of the fused images is verified by means of voting, and the adjacent blocks from different candidate images are weighted and fused. Experimental results prove that this algorithm is faster and more precise than traditional wavelet-based fusion.

Key words: image fusion, multi-focus image, definition, region acutance

中图分类号: 

• TP391.41