Classification of Breast Cancer Using SVM Classifier Technique

This paper proposes a technique for classifying the breast cancer from mammogram. The proposed system aims at developing the visualization tool for detecting the breast cancer and minimizing the scheme of detection. The detection method is organized as follows: (a) Image Enhancement (b) Segmentation (c) Feature extraction (d) Classification using SVM classifier Technique. Image enhancement step concentrates on converting an image to more and better understandable level thereby applying Median filtering approach for reducing the noise on an image. Then the Contrast stretching is applied to increase the contrast of the image. The main part of cancer detection is segmenting the breast image to improve the diagnosis and detection of breast cancer. The segmentation used here is Thresholding. Next the features are extracted from the cancer segmented area and classified the cancer according to its feature by using Support Vector Machine (SVM) classification technique. The method was tested for 119 mammogram image from the mini-mias database.