Brain tumor segmentation and classification using DWT, Gabour wavelet and GLCM

The brain is the central control unit of human body. Tumor affects the brain means it causes the death of the patient if it is not diagnosed in the early stage. Among the different available imaging modalities, Magnetic Resonance Image (MRI) is a better method for area calculation and classification based on the grade of the tumor. MRI doesn't produce any harmful radiation. Nowadays there exists no automatic system to detect and identify the grade of the tumor. This paper proposes a technique to segment and classify the brain tumor from MRI. It helps doctors to plan further treatment or surgery. This paper used Ostu's threshlding for preprocessing. The tumor is detected with the help of K-Means clustering. Discrete wavelet transform(DWT), Gabour wavelet and GLCM features are extracted. The feature set reduction is made possible by Principal component analysis(PCA). Finally the tumor is classified as benign or malignant using support vector classifier(SVM). Linear kernel, polynomial kernel, and RBF kernel were applied and the result shows that liner kernel produces more accuracy.