Automatic Skin Lesion Detection Using DWT Thresholding based Otsu Segmentation Method

Abstract

Skin is exposed to outer environment thus the disease and infection occurs more to skin. Skin cancer is one of the dangerous diseases and the deadliest form is melanoma. There is a need for computerized system to detect skin lesions and evaluate a patient’s risk of melanoma. Segmentation not only plays an important role but also a challenging task in skin lesion detection due to low contrast between lesion and the surrounding skin, the irregular and fuzzy lesion borders and various imaging acquisition conditions. Therefore, an automatic skin lesion segmentation system is proposed in this paper. In segmentation process Discrete Wavelet Transform (DWT) thresholding based on Otsu method is applied to extract lesion areas.The performance of the proposed system is tested on the ISBI 2016: Skin Lesion Analysis Towards Melanoma Detection challenge website and the result is very appreciable and promising accuracy in segmentation process.