MULTIPLE TUMOR & INFECTION DETECTION IN MRI BRAIN IMAGE USING SVM CLASSIFIER

Abstract

A support vector machine (SVM) classifier is being proposed for classification of brain tissues in magnetic resonance photos (MRI). A wavelet based totally texture features set is derived. The most useful texture capabilities are extracted from regular as well as tumor regions through the usage of spatial gray level dependence technique. The proposed technique resolves the massive trouble of category strategies. These most efficient functions are then used to categories the brain tissues into benign and malignant tumor. The overall performance of the set of rules is evaluated totally based on a sequence of brain tumor images.