Abstract—Skin cancer is the rapid growth of abnormal skin tissues. There are several types of skin cancer, and the deadliest being melanoma. The most important warning sign for melanoma is a lesion on the skin that keeps on developing in shape, color or size. Melanoma must be detected in the initial stage, for early medical treatment. In this research, we propose and approach to automatically classify melanoma. Otsu method was used to get the lesion region. A feature that is usually used in melanoma classification in visual checking was use. The additional feature that we also use is adaptive additional shape feature (ADPS) to specific changing of curve. In classification process we use deep neural networks for skin lesions classification. We also compare deep neural networks with support vector machine. The experimental results of classification demonstrated that ADPS with other features using deep neural network as classifier provided good classification in evaluation process.
Index Terms—Image segmentation, feature extraction, melanoma detection, skin cancer.
K. Chaiyakhan is with the Computer Engineering Department, Rajamangala University of Technology Isan, Muang, Nakhon Ratchasima, Thailand (e-mail: kedkarnc@hotmail.com).
P. Chujai is with the Electrical Technology Education Department, Faculty of Industrial Education and Technology, King Mongkut’s University of Technology Thonburi, Bangkok, Thailand (e-mail: pasapitchchujai@gmail.com).
R. Chanklan and N. Kerdprasop are with the School of Computer Engineering, Suranaree University of Technology, Nakhon Ratchasima, Thailand (e-mail: arc_angle@hotmail.com, nittaya.k@gmail.com).
[PDF]
Cite: K. Chaiyakhan, P. Chujai, R. Chanklan, and N. Kerdprasop, "Classification of Melanoma Considering the Adaptive Additional Shape Feature," International Journal of Future Computer and Communication vol. 8, no. 4, pp. 114-118, 2019.
