DocumentCode :
655417
Title :
Performance Analysis of Magnetic Resonance Image Denoising Using Contourlet Transform
Author :
Padmagireeshan, S.J. ; Johnson, R.C. ; Balakrishnan, Arun A. ; Paul, Varghese ; Pillai, A.V. ; Raheem, A. Abdul
Author_Institution :
Dept. of AE&I, Rajagiri Sch. of Eng. & Technol., Kochi, India
fYear :
2013
fDate :
29-31 Aug. 2013
Firstpage :
396
Lastpage :
399
Abstract :
A medical image denoising algorithm using contourlet transform is proposed and the performance of the proposed method is analysed with the existing methods. Noise in magnetic resonance imaging has a Rician distribution and unlike AWGN noise, Rician noise is signal dependent. Separating signal from Rician noise is a tedious task. The proposed approaches were compared with other transform methods such as wavelet thresholding and block DCT. Hard, soft and semi-soft thresholding techniques are described and applied to test images with threshold estimators like universal threshold. The results are compared based on the parameters: PSNR and MSE. Numerical results show that the contour let transform can obtained higher PSNR than wavelet based and block DCT based denoising algorithms.
Keywords :
biomedical MRI; image denoising; medical image processing; source separation; transforms; AWGN noise; MSE; PSNR; Rician distribution; Rician noise; block DCT; contourlet transform; hard thresholding techniques; magnetic resonance image denoising; medical image denoising algorithm; semisoft thresholding techniques; signal separation; soft thresholding techniques; threshold estimators; universal threshold; wavelet thresholding; Discrete cosine transforms; Magnetic resonance imaging; Noise; Noise reduction; Wavelet analysis; Wavelet transforms; Magnetic Resonance Image; contourlet transform; denoising;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Advances in Computing and Communications (ICACC), 2013 Third International Conference on
Conference_Location :
Cochin
Type :
conf
DOI :
10.1109/ICACC.2013.84
Filename :
6686416
Link To Document :
بازگشت