A Novel Image Segmentation Method for Early Detection and Analysis of Melanoma

Abstract

Skin cancer through all types of cancers, is the least frequent and most common form of human cancers. Skin cancer may be Melanoma, Basal and Squamous cell among which Melanoma is the most precarious. It is compulsory to recognize it in its early stage for its proper cure. In this paper, image segmentation method is implemented in MATLAB followed by STOLZ and TDS algorithm.
An image segmentation method is used to classifying the current state of melanoma lesions. It further includes various steps for analysis; pre-processing which is used for image enhancement, removal of noise, hair.
Then the boundary of the skin lesion is detected with the help of initial segmentation. The lesion is investigated to find the parameters using feature extraction property. The fetched parameter values are therefore used to calculate STOLZ algorithm and by feeding the output of the STOLZ algorithm in TDS,
Detection of the current stage of melanoma can be identify.
Keywords: Image segmentation, Feature extraction, STOLZ algorithm, TDS.