Computer vision for diagnosis of malignant melanoma by pixel intensity matrix parameters

Abstract

Malignant melanoma is one of the most frequent types of cancer, and melanoma is the most aggressive type of skin cancer. The malignant melanoma is curable if it is diagnosed at early stages. Main factors that cause melanoma are over exposure to sunlight and genetic factors. Every year 132000 people are affected by skin cancer over the worldwide. Early diagnosis will not only save the life of the people but also the quality of their life. Image processing techniques play a vital role in the diagnosis of malignant melonama. In this paper a novel method is proposed; The correlation factor which is found from the pixel intensity matrix of normal and abnormal mole images are found to make a decision the results obtained for a set of images show that the diagnosis of malignant melanoma by the proposed method is possible.