Segmentation Using Outlier Based Adaptive Thresholding

Image segmentation plays an important role in imageanalysis as a frequent pre-processing step, which divides the imageinto set of different segments. Thresholding is an easy yet efficientmethod for image segmentation, while dividing different objectswith distinct gray levels. Finding an effective threshold is especiallycomplicated task in the segmentation. In this paper, for efficientthreshold selection fuzzy methodology used which produces bettersegmentation results than other methodologies. It was observed thatat different background intensity levels favourable results wereobtained.