Pattern Selective Image Fusion for Multi-focus Image Reconstruction

This paper presents a method for removing focal blur in images using fusion algorithm. Different focal images can be combined to give a single fused image with all focused information. Pattern selective image fusion provides a mechanism for combining multiple images through identifying salient features in the source images and combining those features in to a single fused image. However one of the typical drawbacks of the image fusion algorithm is the reduction in contrast which is undesirable. Loss of contrast in images can be typically attributed to loss of pixel information. The post processed image from fusion algorithm can be further processed using histogram filling algorithm which will lead to better sharper images. Therefore the contribution of this paper is twofold (a) removal of focal blur with the help of effective fusion algorithm, (b) recycling of pixel information to improve contrast and produce better results. Experimental results show the performance of the depth of focus extension using consumer video camera outputs. Keywords—auto focusing, contrast enhancement, histogram filling, focal blur, image fusion.