Computer Vision Based Skin Disease Recognition

Abstract

All over the world, various types of skin diseases are affecting an increasing number of people. It is incredibly simple to diagnose diseases with the aid of modern medical technology. However, there are times when people are unable to quickly travel to a hospital or diagnostic facility and it is also extremely expensive for most people, particularly in developing nations like Bangladesh. It is a good idea to have a backup plan in case something goes wrong, but it's not always possible. This is primarily an image processing method. Then total twelve features have been extracted from our all datasets consist of five diseases (Actinic keratosis, Benign, Eczema, Malignant, Psoriasis). Then, to learn our model, four different machine learning algorithms have utilized. For the psoriasis disease, the highest accuracy from the Random Forest classifier has obtained, which is 88%, and the F1-score is 96%. Therefore, this model can accurately recognize skin diseases with great effectiveness.