Adaptive windowing and granular computing based image segmentation

Abstract

In this paper, adaptive windowing based segmentation scheme has been proposed. The image has been partitioned into different windows and the windows to be segmented, have been fixed by three criteria. The first one is based on pyramid approach where, the preselected windows are merged based on entropy measure. The second one is based on incremental window selection method. In the third criterion, the preselected windows are merged based on entropy measure. The windows thus fixed are considered as sub-images and each sub-image has been segmented based on the notion of rough entropy and granular computing. The algorithm could segment the images with uneven lighting condition.