Malignant melanoma diagnosis using machine learning method based oncombination of nonlinear and texture features

In this article, we have presented a new method for diagnosing skin diseases. The proposed methoddistinguishes between benign and malignant skin diseases (melanoma). The main idea in theproposed method is a new descriptor based on a set of features. In this way, after the images arepre-processed and then blocked, low-level features are extracted from each image block. Thesefeatures include texture and color. Then the bag of features is extracted from the low level features.Bag of words extraction is based on K-means clustering algorithm. The proposed method wasevaluated on a standard data set and the accuracy results of the proposed method are far better thanother previous methods.

Keywords:

skin diseases detection , machine learning , melanoma.