An efficient system based on transforms technique with SVM classifier for diagnosis of tumor in MRI Brain images
Brain Tumour diagnosis is a critical application where lot of emphasis is needed. Conventional image processing constitutes pre-processing, feature extraction and classification mostly in medical imaging image segmentation is popular. In feature extraction, Discrete Wavelet Transform (DWT) and Discrete Fourier Transform(DFT) are used to reduce the coefficient of the image. For Discrete Wavelet Trasform, Haar wavelet is used as a mother wavelet. Three level decomposition are used to represent the feature vector for better classification of MRI brain image. In classification, two different classifiers K Nearest Neighbors (KNN) and Support Vector Machine (SVM) used to classify the tumours in image. sensitivity, specificity, and accuracy are the evaluation metrics to measure the performance of the two classifiers. Comparing the result it is shown that combination of DWT with SVM got the 95.4% accuracy.