Construction of fuzzy edge image using Interval Type II fuzzy set

• Main Authors: Tamalika Chaira, A. K. Ray
• Format: Article
• Language: English
• Published: Atlantis Press 2014-08-01
• Series: International Journal of Computational Intelligence Systems
• Subjects: Type II fuzzy set; fuzzy T norm; fuzzy T co norm; fuzzy divergence; medical image
• Online Access: https://www.atlantis-press.com/article/25868508.pdf

In this paper, a novel method to generate fuzzy edges in medical images using the Type II fuzzy set theory is presented. Medical images are normally poorly illuminated and many edges are not visible properly, so construction of fuzzy edge image is a difficult task. Fuzzy edges are not the binary edges but it signifies the change in intensity levels of the image. The method is based on computation of minimum and maximum values of the intensity levels of the image in a 3x3 pixel neighborhood to form two image matrices with maximum and minimum values. For better representation of uncertainty, Type II fuzzy set is applied to compute upper and lower membership levels of each image matrix. Divergence is computed between the two levels of the maximum value image matrix and also for the minimum value image matrix. Finally, difference between the divergence matrices produces an edge image. Experiment has been performed on several poorly illuminated medical images and the edges are observed to better when compared with the existing edge methods.