Skin Cancer Classification Using CNN


  • Ohood Fahdil Alwan, Dhahir Abdulhade Abdula


In the world wide mostly of the  cancer are skin cancer, essentially, there are two kinds  of skin cancer called malignance and non-malignance.In this paper , design a system to detection and classification skin cancer with high accuracy and sensitivity by using Convolution Neural Network (CNN) which able to diagnose different types of cancer in a human skin. With the objective of utilizing more meaningful information to improve skin cancer and help physicians in the clinical diagnosis and accurate detection of the disease. The system supports physicians to prevent errors while identifying and classifying cancer. This is motivated by potential performance improvement in the general automatic and giving reliability in decision-making and rapid detection of skin cancer this technology is of great and economic importance to physicians.The system divided in two type which contain the following stages: image acquisition, preprocessing, and classification, while the second part consist of image acquisition, classification.There is a significant change between the classification with preprocessing and without preprocessing, as with preprocessing the accuracy decreased that return to the reason that the pictures that were taken to the skin are too close and the do not require any preprocessing.

Keyword: Skin , Cancer , Algorithm CNN