A New Modified Kohenon Algorithm for Automatic Skin Cancer Detection
6 Pages Posted: 7 Aug 2019
Date Written: August 2, 2019
Skin illnesses are tumors that due to the improvement of uncommon cells which could assault or spread to other parts of the body. There are three essential composes basal-cell tumor, squamous-cell tumor, and cancer. Among the three melanoma spreads through metastasis, and along these lines, it has been becoming out to be extraordinarily lethal. Melanomas, for the most part, happen inside the pores and skin what is additional, unmistakable proof of skin tumor ought to be practical in perspective on the Melanoma pix. The factor of this test paper is to recollect and discuss the distinctive portrayal of estimations connected to different types of therapeutic datasets and thinks around its execution. The association calculations with the greatest exactness are on exceptional varieties of medicinal datasets are taken for execution investigation. In our proposed work, modified Kohonen neural community is used for recreation plan reason, which bunches the given informational series into carcinogenic and non-threatening. The Kohonen self – organizing map is one of the neural machine unsupervised studying algorithms. This calculation is utilized as a part of looking after the issue in extraordinary areas. This proposed technique includes preprocessing, postprocessing, segmentation curvelet domain Feature Extraction, and classifiers. Intype several methods including Support vector system, Three layers neural Network, Back Propagation Neural Network, Modified Kohonen networkself-organizing map) are employed to separate the skin cancer types and their Accuracy performance equated using the various parameters.
Keywords: Modified Kohonen Network, Curvelet Transform Segmentation, Noise Removal, Feature Extraction
Suggested Citation: Suggested Citation