Skin Disease Detection Using Image Processing and Neural Networks

Abstract

Dermatology is one of the diseases which is unpredictable and difficult to diagnose because of its complexity. Dermatology is very difficult to diagnose because of its complexity and expensive. In most developing countries, it is expensive for people to detect. According to world Health Organization (WHO), skin disease is the most common in India. The present use of smartphones in the developing countries like India has opened new avenues for diagnosis of disease inexpensively. In smartphones camera technology can be used to exploit the image processing capabilities of the device for disease diagnosis. The proposed system deals with the creation of the application that helps to diagnose the skin disease. This application uses machine learning and image processing technologies to detect the type of the skin disease. The system consists of 2 parts- machine learning and image processing. Image processing part deals with applying the filters to images to remove noise to make it uniform. Unwanted elements must be removed from the image before processing. If unwanted elements are not removed it will affect the output efficiency. The machine learning part deals with the processing of data from the image and generation of the result.