Image thresholding by minimisation of fuzzy compactness and linear index of fuzziness

This paper describes a study of two fuzzy based thresholding algorithms for digitising images using the index of fuzziness and fuzzy compactness. The problem of histogram sharpening and thresholding by minimizing grey level fuzziness and compactness is considered. The choices of appropriate membership functions and the optimum value of its bandwidth is seen to be restricted enabling the methods more flexible. The effectiveness of the two methods is demonstrated for different bandwidths of the membership function using ill-defined images (letter and medical images).