Multifocus image fusion based on region selection

Abstract

In this paper, a novel multifocus image fusion method based on region selection is proposed. The basic idea is to select sharply focused regions from source images and to combine them together to reconstruct the resultant image. Region Acutance as a kind of superior sharpness criterion is introduced here for region segmentation. In view of the inevitably errors of selection, morphological operations are used to eliminate the discontinuity in the interior of regions. Experimental results demonstrate that the proposed method outperforms traditional fusion methods both in visual effect and objective evaluation.