Open Access Open Access  Restricted Access Subscription or Fee Access

Classification of Mammogram Images Using Multi-SVM

A. Anbumani, Suresh Kumar P., Sathishkumar A.


Bosom malignant growth is normal in ladies these days. It first starts when cells in the bosom start to develop wild. These cells for the most part structure a tumor that will frequently be seen on an x-beam or felt as an irregularity. Cells in about any piece of the body can become malignant in growth and can spread to different regions of the body. In the current framework, rule based methodology is utilized in order which gives a static range an incentive for various classes. Along these lines we won't capable, powerful pictures or exception conduct pictures. Highlights set is not standardized. Thusly various highlights show various yields and show diverse portrayal during preparing period of classifiers. Classifiers are not ready to recognize include covering. Hence at the learning stage, an example of the picture is not recognized. The multi - characterization issue is not streamlined and disregard the class unevenness issue. In this way, in adapting all classes is not contributing to learning stage so it turned into a one-sided learning. The proposed methodology first starts when cells in the chest begin to create insane. These cells by and large structure a tumor that will normally be seen on an x-shaft or felt as a projection. Cells in about any bit of the body can advance towards turning out to be a threat and can spread to various zones of the body. There are just around six periods of chest threat. It is always found that the revelation of threatening development at the chief stage can fix it. A model picture is taken as an information and differentiated and the photos recently set away in the database related to harm. If the disclosure is found productive, by then, relating treatment is proposed. The period of harm is being appeared and separate treatment is being urged to the patient. Stage keen treatment and prescriptions are given to fix that dangerous development.

Keywords: Fuzzy C means clustering, Gaussian filter, GLCM, Haar wavelet transform, image scaling, morphological, region of interest, statistical features, K-implies clustering, multi-SVM (Support Vector Machine)

Cite this Article Anbumani A., Suresh Kumar P., Sathishkumar A. Classification of Mammogram Images Using Multi-SVM. Journal of Image Processing & Pattern Recognition Progress. 2020; 7(1): 32–51p.

Full Text:




  • There are currently no refbacks.