Multi-focus Image Fusion Based on Side Window Filtering Technique and Majority Filter

Abstract

Multi-focus image fusion is applied in the computer vision widely which can increase depth-of-field from multi-focus photograph. In this paper, we present a novel multi-focus image fusion method based on the side window filtering (SWF) technique and majority filter. The method effectively improves the edge preserving capability using the SWF technique to extract the features of the source images well. Firstly, to generate the accurate focus map, the rough focus map obtained by mean filter is refined by the SWF technique in focus region detection. Then the initial decision map is obtained using the pixel-wise maximum rule, and then the initial decision map is regularized and optimized by majority filter and the SWF technique to get the final decision map. Finally, the all-in-focus image is reconstructed by the final decision map. Experimental results demonstrate that the proposed method has better performances than other fusion methods in visual effect and objective evaluation.