AOSTKACTA concept or correlation between twO properties (fuzzy repre~cnlations) of an image IS introduced. A sel of algorithms for image segmenLation (both fuvy and nonfuzzy) has been formulated. The spatial infonnation is taken care of by the following measures: transitional correlation and within-class correlation, A relation between the correlation coefficient and thc inde~ of fuzziness is theoretically eSLablished and e~perimenlally verified, The effective ness of the algorithms is illustrated on images having different types of hiSlOgrams.