14 – Region-based multi-focus image fusion

Recently, region-based image fusion has attracted considerable attention because of its perceived advantages, which include: (1) The fusion rules are based on combining regions instead of pixels. Thus, more useful tests for choosing proper regions from the source images, based on various properties of a region, can be implemented prior to fusion. (2) Processing semantic regions rather than at individual pixels can help overcome some of the problems with pixel-fusion methods such as sensitivity to noise, blurring effects and misregistration [1].