Brain Tumor Segmentation with Wavelet Watershed and Detection Using Multi-SVM Classifier

Abstract

Detection of tumour in brain has most prominence in the recent years. Various processes are proposed for detecting BRAIN TUMOUR which comprises with image segmentation and classification process. But classification process has dominant and suppressed most of the techniques by its advantages of detecting and classifying brain tumour. In this paper a novel approach of Wavelet watershed technique is proposed with MULTI RBF SVM classifier process for segmentation and classification processes respectively. The feature extraction and region segmentation processes were completed by Wavelet Watershed technique for this we used to calculate the energy of the image for a texture level classification mode. Under Multi SVM classifier the weight comes into play for training datasets along with classification mode. Experimental results are acquired from the proposed technique is about 95%.