“…In the past few decades, a variety of methods, such as statistical region merging (SRM), dynamic region merging (DRM), split and merge, similarity based merging, hierarchical merging, boundary based merging etc., have been developed to eliminate the segmentation errors [10][11][12][13][14][15].Here, image regions can be described in terms of shape, size, color, texture etc., and utilization of these kinds of parameters in region based merging techniques will enhance the efficiency of the method and improve the segmentation patterns [17][18][19][20], which are purely based on probabilistic and some logical criteria. In case, to determine the similarity between the two regions depends on pixel mean, variance and gray level, also which is based on Euclidean distance and log-likelihood ratio are wellknown statistics for region evaluation.…”