Purpose <p>Pore and skin cancer is one of the most harmful cancers. It should be diagnosed and treated early; to avoid unnecessary emergencies.</p> Methods <p>For early detection, a reliable automated machine for the popularity of pores and skin lesions is practically mandatory in order to reduce human effort, time and lifestyle. In this research, Keras prepossessing is used to describe the boundaries and edges of images and convolution neural networks are used.</p> Results <p>CNN can perform skin cancer classification with a higher level of accuracy and efficiency than existing methods. The system’s core is in the way it applies cancer research know-how and public health databases to perform its work.</p> Conclusions <p>This work develops a model that can assist in determining skin cancer, and the accuracy was above 87%.</p>