Melanocytic Nevus Reduction Algorithm Using RGB Channels' Modes

Abstract

© 2019 IEEE. There are several pleasant and effective systems or algorithms regarding humans' mole. All of these system pay attention to the field of health and medicine. This study focuses on physical aspect of moles in images. In this study, the researchers use digital image analysis to design an algorithm and implement a system to detect a mole in an images and then remove the mole thorough converting the color of the mole. The system contains two parts. In the first part, the system detects the mole following different steps such as grayscaling using average method, Gaussian blurring, binarization using Otsu method and contour detection. In the second part, the system performs mole removal in which system replaces the color of mole's pixels to a color that is made by modes in different channels. The researchers tested the system regarding mole detection and mole removal separately and the results are presented in different tables. Based on the results, the accuracy of the system regarding mole detection is 90% while regarding mole removal is 85%.

Source or Periodical Title

Proceedings of 2019 International Conference on Computational Intelligence and Knowledge Economy, ICCIKE 2019

Page

98-103

Document Type

Article

Subject

average method, digital image analysis, Gaussian blurring, mode, mole, skin

This document is currently not available here.

Share

COinS