Skin Tumor Segmentation Using Artificial Neural Network in Ultrasonic Images

Abstract

One of the world's most popular and rising health conditions is the skin. The human skin tumour, because of the nuances in texture, colour, hair appearance and other attributes, is the most volatile and one of the toughest organisms to immediately identify and determine. In this project we have proposed a device that uses Artificial Neural Network to identify skin tumours. Different forms of dermatological tumours are effectively identified. It primarily comprises of three steps of picture preparation, teaching and identification. We add algorithms including a transition from grey to HSV to the input picture during the image processing step. The input image is observed with artificial neural network algorithms after HSV values are collected. The percentage of contamination is often identified as an exception to the identification.