Classification algorithms for effective detection and prevention of the melanoma and benign skin lesion

Abstract

Cancer is a serious dead disease today that is affected by a division which is unable to control the abnormal cells in the human body Hence the rapid growth of cancerous is able to invade and to destroy the adjacent structures and also spread in to the distant sites to cause death Moreover the cancer is also able to affect all the living cells of all levels of human bodies without gender wise Nowadays the skin cancer is growing the skin cells enormously and it is also appears while the process of unrepaired DNA damage to the skin cells mutations In addition the skin cancer is also classified into three different kinds of skin cancers like basal cell squamous and melanomas skin cancers Even though the major classifications of skin cancer are melanoma and non melanoma Here the Melanoma is a malignancy of the cells which provide the melanocytes skin colour and able to invade the neighbouring tissues In addition it is also able to spread throughout the human body and it may lead to cause of death On the other hand the non melanoma spreads in to other parts of the body For saving the human life and protect the people from the disease skin cancer this research work introduces new prediction system which is able to predict the skin cancer in advance based on the occurring symptoms in human body The proposed prediction model carried out three major tasks namely data pre processing clustering and the classification First new pre processing activities are considered like image conversion contour detection segmentation wavelet transform and feature extraction has been carried Here new segmentation and feature extraction methods also have been proposed for effective pre processing newline