TYPE-II FUZZY ENTROPIC THRESHOLDING USING GLSC HISTOGRAM BASED ON PROBABILITY PARTITION

Abstract

Image segmentation of threshold based is a useful technique in the preprocessing phase of image processing applications. Some  two dimensional entropy, methods use  local properties  of the  image to compute the optimal threshold.  Yang Xiao et al. simplification on this procedure worked well with the inclusion of spatial correlation features which reduces the time complexity of earlier methods. Seetharama Prasad et al. improvised further in the process of obtaining the varying similarity measure. In this paper type-II fuzzy membership degrees of gray values are employed with probability partition of the image as object and background probabilities. Spatial correlation parameters are used in the computation of entropy criterion function to obtain   optimal threshold of the image. For low contrast images contrast enhancement is assumed. Experimental results are so encouraging.