In worldwide, breast cancer is considered as the main primary cause of death for the woman. In the diagnosis of the disease, the early detection of malignancy can be increased the survival for women. Various techniques have been introduced for the detection of breast cancer, in which the scheme of mammography has been considered as the most promising scheme and radiologist used this scheme frequently. Generally, in breast mammography, the Mammogram images are represent as low contrast with high noisy, in which the bright region are represented as cancer. So, the cancerous lesions detection has been considered as an active research area in mammogram image. Several schemes has been presented for cancerous detection in earlier, but no one gave best results and not able to satisfy the detection criteria. Still, the early detection and classification of breast cancer has been considered as a big issue. The mammography images with detection and diagnosis process are cost effective and better chances of recovery. In this research, two classification schemes like K-Nearest Neighbour(KNN) and Support vector Machine (SVM) are introduced to segment the breast cancer and classify the disease like begin. Here, the impulsive noise is removed from mammography image to improve the accuracy of classification by using wiener filter. After that, the cancer malignant is segmented by using k-means clustering to reduce the time consumption. Then, the Grey Level Co-occurrence Matrix (GLCM) features are extracted for efficient classification. Finally, the two classification schemes has been classified the cancer disease. After that, the performance of those two methods is evaluated and their performance is measured in terms of accuracy, sensitivity, specificity and F-measure. Presented schemes are validated on Mini-MIAS database and results were compared by using MATLAB.
