Brain CT-Scan Images Classification Using PCA, Wavelet Transform and K-NN

109 All Rights Reserved © 2012 IJARCSEE Abstract: With rapid development of technology in biomedical image processing, classification of tissues of human body is very challenging task as it requires very accurate results without any misclassification. By making use of this technology along with neural network; a hybrid technique has been proposed for classification of Brain CT-Scan images. This technique is not limited to medical field; it is also applicable to classification of natural images. Database consists of CT-Scan images and Brodatz texture. The methodology adopted in this paper consists of two stages: firstly, features are extracted from given images using feature extraction algorithms PCA and Wavelet Transform. They are further fed as an input to train the K-NN classifier to classify between normal and abnormal images. For Brain CT-Scan images; features extracted by PCA gives 100% classification accuracy with execution time of 0.6133 seconds whereas for Brodatz texture images; features by Wavelet transform gives classification accuracy of 100% with execution time of 0.1912 seconds. Code is developed by using MATLAB 2011a.