Multilevel Thresholding Based on Fuzzy Masi Entropy

Multilevel thresholding is one of the convenient and effective methods in image processing. In this paper, we propose a new multilevel thresholding method based on fuzzy Masi entropy to deal with the selection of objective function in multilevel thresholding. The performance of the proposed method is tested by randomly selecting images from berkeley segmentation dataset (BSD-500). Peak signal-to-noise ratio, structural similarity index metric and feature similarity index metric are computed as the indicators to evaluate image quality in various aspects. The segmentation results show that fuzzy Masi entropy as the objective function have higher quality and stable performance than the other compared objective functions.