Detection And Classification Technique Of Breast Cancer Using Multi Kernal SVM Classifier Approach

Abstract

Today, one of the mostly seen disease in women is Breast Cancer. It is circulated different countries of all over the world. Mammography is a kind of low-powered X-ray diagnosis approach for detection and diagnosis of cancer diseases early. This task is done for classification of diseases as Malignant or Benign. The entire work focuses on the basis of two cases. One is detection of different type of tumors as suspicious regions and another is process to extract features from mammogram images and classification of type of tumors presented. There are some phases of detection of tumour: image pre-processing, image enhancement using histogram, extraction of features from mammographic images, Segmentation using Otsu thresholding method, classification using Support Vector Machine (SVM) classifier. Image Preprocessing is basically done by applying two dimensional median filter and histogram equalization for getting more enhanced image. Then extraction of features set is performed from the images. Here the different types of tumor like Benign, Malignant, or Normal image are classified using the SVM classifier. In this technique, we have used statistical parameter like as entropy, mean, RMS, correlation, variance, standard deviation. This techniques shows how easily we can detect region of tumor is present in mammogram images with more than 80% of accuracy rates for linear classification using SVM.