Microscopic Image Thresholding using Restricted Equivalence Function based Fuzzy Entropy Minimization and Bat Algorithm

Ubiquitous use of microscope in the field of medical diagnosis influences the development of automated systems. The inbuilt noise, illumination and contrast variations make microscopic image processing an emerging field of computer vision applications. This paper presents a novel and fast thresholding technique for microscopic data. To represent the inherent image vagueness, we use a fast processing fuzzy membership value generation technique using restricted equivalence function (REF). Then, a fuzzy entropy value is used to measure the total fuzziness present in the object and the background of the image. Finally, to search the optimal threshold value we use the well popular Bat algorithm. We have also implemented a multilevel thresholding technique for processing some complex fluorescence microscopy images. Experimental results on microscopic data and also on normal images show the superiority of the proposed thresholding technique. Experimental results on microscopic data and also on normal images show the superiority of the proposed thresholding technique. Proposed method is superior than other state-of-the-art methods not only in processing time but also in quantitative results. The proposed method is superior than other state-of-the-art methods not only in processing time, but also in quantitative results.