Expert Skin Disease Identification System Using Machine Learning

Abstract

Dermatology disorders are illnesses that are most prevalent worldwide. While it's that, it is extremely hard to diagnose and needs a hugepractical experience. We are offering an approach in this project to identify multiple diseases. The system has a two stage approach that incorporates computer vision and machine learning on clinically validated histopathological attributes to exactly classify the disease. Our project goal is to diagnose skin disease type easily with accuracy and recommend the latest and most global medical suggestions. In the first step, the skin disorder image is subjected to various pre processing techniques and extraction of features (histopathological attribute). The second stage includes the use of Machine Learning algorithms to find disease that are assisted by the histopathological attributes found on skin examination. This paper proposes an illness of the uploaded image by using Image Processing technique assisted by skin identification system. As an input to the prototype, the person can upload an image of the infected skin by capturing using the Smartphone camera. Then the preprocessing technique is applied on the image to identify the disease and the detected disease is displayed on the screen. This system is very much beneficial for the people who are living in rural areas where the access of dermatologists is difficult. We use Pycharm-based python script for the experimental results for this proposed program.