Description

Skin cancers are cancers that arise from the skin. They are due to the development of abnormal cells that have the ability to invade or spread to other parts of the body. The features were extracted from the input images based on texture or shape based approaches. Inorder to identify both melanocytic skin lesions (MSLs) and non-melanocytic skin lesions (NoMSLs) methods are proposes a new computer-aided method for the skin lesion classification applications. For the skin cancer disease detection, the computer aided base machine learning classification has become the more attention. According to the categorizing of MSL, several methods are implemented to classify the melanoma and nevus. Two types of classification models in this process, First one is a layered model that uses a task decomposition strategy and then the second one is flat models to serve as performance baselines. The performance of the process is measure based on the performance metrics. The process of extraction of three different types of features helps in the exact identification of the features from the images. The process is suitable for all type of images and does not requires a large amount of training set. The performance of the process is improved comparing to the existing methods for the classification of melanoma. The process includes the classification of different types of Melanoma. The images were obtained from the melanoma dataset. The images were segmented based on thresholding process and morphological operations inorder to identify tumor portions more clearly. Three types of methods were used for the extraction of features from the color images. The methods used were color based, sub-region based and texture based features. The extracted features were classified based on the Flat and Layer model classifier. The regions that were classified as tumor portions were represented in separate notations. The performance of the process is measured based on the performance metrics.