ABSTRACT
Skin diseases are quite challenging to diagnose if there is a lack of experience in the field of dermatology due to their complexity and unpredictability. It often requires extensive tests to verify the skin condition that the patient may be facing, which also adds to the duration of the diagnosis and the varying amount of time the practitioners may take. So, we have developed a skin lesions diagnosis system that utilizes complex computational techniques to accurately diagnose the skin lesions and ensure early detection of the skin disease. The system also provides information about skin disease, including its features, common causes, and commonly prescribed treatment, which may be helpful for patients until they can receive medical expertise. The system can accept skin images and process them to enhance the images and remove noise. Features from the images are then extracted using the convolutional neural network (CNN) technique, followed by the classification of the skin disease using the Softmax Classifier's algorithm. The system can classify the skin disease within a few seconds with an accuracy of more than 95%. The system will not only be helpful for self-diagnosis but also medical students in the field of dermatology.
