Transition Region Extraction Algorithm Based on LocalFuzzy Entropy

To improve the performances of transition region extraction algorithmsundernoises, a novel local fuzzy entropy-based transition region algorithm was proposed by making improvement to the existing local entropy-based transition region algorithm. The proposed algorithm makes full use of the difference between transition regions and target regions (or background regions). The performance of the proposed algorithm was compared with the local entropy algorithm underGaussian noise and salt and peppernoise. The comparison shows that the proposed algorithm provideshigheraccuracy than the local entropy-based transition region extraction algorithm does under Gaussian noise, although both behavewellunder salt and peppernoise.