Multi-focus Image Fusion Based on the Filtering Techniques and Block Consistency Verification

In this paper, a novel simple and effective multi-focus image fusion technique is proposed. In the decision maps learning stage, based on the block consistency verification (BCV) and guided filtering techniques, a series of binary decision maps are obtained. In the fusion stage, neighbor distance (ND) filtering are performed on source images, then the informative highpass images and the energetic lowpass images are generated. The fused results are developed by constructing weight maps and the ND filtered images. Compared with three traditional methods and five state-of-the-art fusion methods, experimental results clearly demonstrate the superiority of the proposed method in terms of both subjective assessment and quantitative evaluations.