An Adaptive Thresholds Method for Gray-level Image Segmentation Using Type-II Fuzzy Sets

摘要

Aiming at threshold uncertainty caused by fuzziness ofimage for image segmentation, an adaptive thresholding methodfor gray-level image segmentation using type-II fuzzy sets isproposed . Fuzzy index is got using type-II fuzzy sets techniqe,that can overcome effectively the uncertainty, histogram peaksvalue of an image can be fended out automaticly based onmaximum Mahalanobis distance and minimum classificationinaccuracy rate, and the number of pixel classes is used to controlthe shape of fuzzy index courve of the image , fuzzy membershipfunction window-width can be also got automaticly, and optimumsegmentation thresholds are determined automaticly according tothe fuzzy index courve of the image. Experiment shows that theproposed method is effective for gray-level images.