No Detail Found!

Features Reductions using Color Moments and Classification of Brain MRI using K-NN.

Author(s)
Z. Ullah Computer Engineering Department, Changwon National University, South Korea
S-H. Lee Computer Engineering Department, Changwon National University, South Korea
M. N. Khan Computer Science Department, SZABIST, Islamabad, Pakistan
M. Fayaz Computer Engineering Department, Jeju National University, Jejusi, Jeju Special Self-Governing Province, Korea
M. M. Iqbal Computer Science Department, SZABIST, Islamabad, Pakistan
Abstract
The paper presents an intelligent method for classification of brain Magnetic Resonance Imaging (MRI). Thep roposed methodology comprises of four stages; pre-processing, feature extraction, feature reduction, and classification. In the first stage, a median filter has been applied on brain MRI to remove the noise and then the image converted to RGB. In the second stage, features of the image have been extracted using Discrete Wavelet Transform (DWT). However, the number of extracted features were very high, which have been reduced in the feature reduction stage using color moments. The feature reduction is the main contribution of this paper. Finally, the reduced features are classified into normal and abnormal MRI using kNearest Neighbors (k-NN).The system has proved better results as compared to the other methods with a reduced number of features. The overall accuracy of the proposed method is 94.9745%.
Publication Details
Page(s) 77-83
DOI DOI not available
Published Journal: Technical Journal, Volume: 23, Issue: 4, Year: 2018

Request the full text from PASTIC

Keywords
Principal Component Analysis (PCA), Color Moments (CM), k-Nearest Neighbors (k-NN), Discrete Wavelet Transform (DWT), Approximation Component
0

Citations

0

Downloads

33

Views