Project ID: 229
NORHAMIRA BINTI MOHAMAD - CS230
2017412226
Supervisor: RAJESWARI A/P RAJU
Examiner: SHARIFAH NURULHIKMAH BINTI SYED YASIN
SKIN CANCER CLASSIFICATION USING ARTIFICIAL NEURAL NETWORK
Abstract
Skin cancer is the most common form of cancers in humans. Most type of cancer can be treated at early stages. Early detection is important in order to save the patients. There are two types of skin cancer in this system which is melanoma and carcinoma. Melanoma is the most dangerous skin cancer because it can lead to death. Carcinoma is usually seldom spread to other parts of body. The method used for this system is by using image processing technique that require dermoscopic images and the algorithm that use is artificial neural network. The dermoscopic images of skin cancer is pre-processed for noise removal. Then, the parameters used for feature extraction is contrast, correlation, homogeneity and energy. These features are given as the input nodes to the neural network.