Comparative Analysis of Thresholding Methods to Facilitate Decision Making over Microscopic Skin Image

It is a useful technique used in many image processing applications to distinguish objects from the background. In this work, we examine the thresholding techniques that may be applied to make microscopic skin images easier to grasp in terms of collagen intensity, which can disclose the majority of skin illnesses. On the provided data set, 15 global thresholding techniques that are commonly recommended in the literature were looked at and tried. 10 histopathology images stained with Masson's Trichrome were used to build the dataset. The data set was used to evaluate several thresholding techniques, and ME was computed to choose the best one. The experiments revealed that the Otsu and IsoData approaches produced the best outcomes, with a ME value of 0.063 and 0.073 respectively.