Abstract

The cancer of the skin is a frequent type of cancer, and the chance of survival rises with early identification. One of the deadliest forms of cancer and a major global cause of death is skin cancer. Early detection of skin cancer can help lower the death toll. The most popular, albeit less accurate, technique for diagnosing cancer of the skin is visual examination. There have been suggestions for deep learning-based techniques to help physicians diagnose skin malignancies accurately and early. Early diagnosis of skin cancer signs is imperative due to the disease's rising incidence, high death rate, and cost of care. Given the gravity of these problems, scientists have created a number of early skin cancer screening methods. To develop deep learning models that use the Mobile-net technique to classify skin cancer. Our approach involves utilizing the ISIC dataset, which has 2991 photos of each of the six different kinds of skin lesion cancer, to identify as well as diagnose cancer of the skin using the Mobile-net algorithm. We employ five distinct modules in our article to carry out our project: dataset collection, preprocessing, training and testing data splitting, deep learning algorithm model implementation, and classification and prediction as the last module. Deep learning and image processing concepts are used in the diagnosis process. Based on the findings, the suggested approach performed admirably for a range of skin disorders and also used the patient's metadata together with the disease image for the classification of skin.