Skin Cancer Segmentation with Entropy PAL MCET using Gaussian Distribution

Skin cancer is the most common cancer diagnosis and it is the most preventable cancer. Diagnosis of skin cancer would be improved if an accurate skin image segmentation is available. The process of image segmentation is a fundamental step in many applications of image processing, yet current methods and techniques for image segmentation necessitate particular domain knowledge to define well the region of the cancer. To estimate an optimal threshold for skin cancer images, thresholding is used as the principal approach of segmentation. We propose a new method for skin cancer segmentation using a Minimum Cross Entropy Thresholding (MCET) method. We applied this method on bimodal skin cancer images and obtained promising experimental results. The resulting segmented skin cancer images yielded better estimation of the optimal threshold than does the same MCET method with Poisson distribution.