Skin cancer is a dangerous disease. Benign and malignant melanomas are one of the skin cancer diseases. Melanoma is a highly dangerous disease. It can be curable if it is detected early. Benign can be cured easily but malignant cannot be cured fastly. Benign and malignant melanoma appears in the early stages while differentiating them. Different methods have been used for differentiating them. Skin cancer can be detected in early stages by visualizing with clinical screening by dermoscopic analysis. Detecting automatically skin lesion is a typical task. Skin cancer symptoms are small blood vessels visible, thickened patch, ulcers and bleed. Skin cancer detected by capturing images with a skin magnifier with polarized light and diagnosed with deep learning classifier in which data augmentation and weights can be added to it. In this CNN classifier is used in which RESNET-50 and VGG-16 were used in which image were resized and weights were added and then the augmentation of the data can be done.