An Efficient WOA BasedDeep Neural Network for Skin Disease Classification

Abstract

Dermoscopy is an imaging technique to diagnose the skin diseases at the early stage. Dermatological disease can be diagnosed by a dermatology specialist through the interpretation of medical dermoscopy images by machine learning. The approachesexistto identify the skin diseases does not provide better prediction rate due to the problem of low contrast and blurring. In our paper, we propose a skin disease identification approach for effectively identifying the diseases. Initially Level set Model (LSM) is used for image segmentation. Then feature extraction is processed for extracting the important features. Deep Neural Network (DNN) based Whale Optimization algorithm (WOA) is used for identification of skin diseases. The proposed model delivered 92% accuracy in skin disease identification and proved to be the best automatic model for skin diseases.