Automatic Triple-A Segmentation of Skin Cancer Images based on Histogram Classification

Abstract

Skin cancer has been the most common and represents 50% of all new cancers detected each year. If detected at an early stage, simple and economic treatment can cure it mostly. Accurate skin lesion segmentation is critical in automated early diagnosis system. This paper present a triple segmentation procedure based on the pixels distribution Bell-shaped (Normal), J-shaped, Reverse J-shaped and U-shaped peaks that is bimodal. According to the nature of dermoscopy images distributions, three segmentation methods are used to identify the normal skin cancer from malignant skin and to extract the tumor region. First, active contours are used for bell distribution shape. Second segmentation is done using adjusted ant colony optimization when the U-shaped peaks distribution was classify. Third segmentation strategies apply adaptive threshold for two J-shapes. Experiments on synthetic and real dermoscopy images demonstrate the advantages of the proposed methods that is able to produce ant colony optimization accurate segmentation when applied to a large number of skin cancer (melanoma) images.