Skin Cancer Image Segmentation using Convolutional Neural Network with Tensorflow
Author(s)
Abstract
Abstract— Skin cancer is one of the most common cancers worldwide, and its occurrence is increasing rapidly due to excessive exposure to ultraviolet (UV) radiation, environmental changes, and unhealthy lifestyles. Early detection is very important because it improves treatment success and survival rates. Traditional diagnosis methods such as dermoscopy and biopsy are costly, time-consuming, and require expert doctors. To overcome these challenges, this project proposes a web-based skin cancer detection system using deep learning techniques. The system is developed using Flask as the backend framework and uses a Convolutional Neural Network (CNN) for image classification. TensorFlow and Keras are used to train the model with dermoscopic skin images. Users can register, log in, and upload skin images through a simple interface. The uploaded image is preprocessed and analysed by the CNN model to predict the type of skin cancer in real time. The system also provides basic treatment suggestions to create awareness among users. MySQL is used to securely store user and login details. An admin module is included to manage user information efficiently. Additionally, a chatbot feature is integrated to answer user queries instantly. The proposed system provides a fast, reliable, and accessible method for preliminary skin cancer screening. Although it does not replace professional medical diagnosis, it acts as a supportive tool for early detection and awareness. By combining artificial intelligence with web technologies, the system improves accuracy, efficiency, and accessibility in skin cancer detection
Keywords