An Automated Dermatological Images Segmentation Based on a New Hybrid Intelligent ACO-GA Algorithm and Diseases Identification Using TSVM Classifier

Dermatological diseases are the common and most extensive diseases worldwide. Diagnosis of dermatological diseases as well as skin cancer diseases is extremely difficult during its early to mid-stages by visual examination. Current researches propose efficient approach which can identify maximum 3- 9 types of skin diseases. It is necessary to develop an automated system which can identify more than 9 types of skin diseases. In this paper, we propose a new automated system for various kinds of dermatological diseases detection. In this automated system, a novel hybrid intelligent ACO-GA algorithm including ACO Algorithm, GA and tabu list is used for different kinds of skin lesion segmentation of dermatological images and Transductive Support Vector Machine (TSVM) is used for dermatological diseases identification. Ant Colony Optimization Algorithm and Genetic Algorithm work simultaneously to inquire most optimal cluster centers from the problem set. After image segmentation, pixels are divided into different classes. In testing phase, the testing samples are compared with training samples using TSVM classifier and features extraction which can give the accurate disease detection result. The result of the proposed system successfully finds out 24 different patterns of dermatological diseases with an accuracy rate of 95%.