Skin Cancer Detection using Convolutional Neural Network Models: VGG-16, VGG-19, ResNet and DenseNet

Skin cancer is a major type of cancer that begins in the skin cells. It occurs when abnormal skin cells grows out of control, usually triggered by exposure to ultraviolet radiating from the sun or for other reasons. Early detection and treatment are critical for favorable skin cancer outcomes. Regular self-examinations of the skin, in addition to expert skin exams, can aid in the detection of worrisome lesions. The technique for successful skin cancer treatment employs both image process and deep learning. The research proposed a technique for classify the skin cancer types. In this work, two forms of skin cancer were finding: benign and malignant. Convolutional Neural Network Models' performance and capability are evaluated. The goal is to create a system model that identify and categorized the skin cancer using Convolution Neural Network Models such as VGG-16, VGG-19, ResNet, and DenseNet. The diagnosis methodology makes use of the image process and deep learning techniques. The number of skin images has been increased by employing different image enhancement methods. Finally, the transfer approach is applied to improve the classification category accuracy even further. VGG-16, VGG-19, ResNet, and DenseNet have performance accuracy of 72.61, 73.19, 84.35, and 82.37 percent, respectively.