One of the most common types and the most lethal form for cancer of skin are the Melanoma and Basal Cell Carcinoma. Early detection of melanoma that too in early stages is always treatable with minor surgery. Effort has been taken to study melanoma systematically in the project. Melanoma is studied using wider, deeper, and highresolution images with deep CNN techniques that provide better performance and accurate results. The dataset collected is augmented using suitable transformations and the system is evaluated using a variety of metrics. The over-elaborate skin cell’s features are extracted using a fitting feature extraction technique and are segmented. We use several models in this implementation, and the results are analyzed based off of the metrics. The model that performs the best is chosen as the best performing model. The proposed system is tested on a large dataset to obtain the best results. Based on metrics, we compare the different models used in this application and find the most efficient one. The new system is put to the test on a huge database, and the results are received.