RSIS International

Comparative Evaluation of Rician Noise Denoising Techniques for MRI Images

Submission Deadline: 30th October 2024
October 2024 Issue : Publication Fee: 30$ USD Submit Now
Submission Deadline: 19th October 2024
Special Issue on Education, Economics, Management, Public Health: Publication Fee: 30$ USD Submit Now

International Journal of Research and Scientific Innovation (IJRSI) | Volume V, Issue II, February 2018 | ISSN 2321–2705

“Comparative Evaluation of Rician Noise Denoising Techniques for MRI Images”

1Saylee Lad, 2Dr. Prof M. S. Panse

IJRISS Call for paper

 1M.Tech Electronics, 2Professor
Electrical Engineering Dept., VJTI, Mumbai, India

Abstract—Image processing plays an important role in Brain tumor detection. MRI images are usually preferred for brain tumor analysis. These images are highly affected by Rician noise. Images should be noise free because noisy image analysis results in poor accuracy. Comparative evaluation of denoising techniques is carried out where the Median filter, Weiner filter, Hard Wavelet Transform and Soft Wavelet Transform are used. Filter performance is evaluated by Mean Square Error and Peak Signal to Noise Ratio. Hard Wavelet Transform is proved to be the best filtering technique for Rician noise removal.

Index Terms— Brain Tumor, Image processing, MRI, Filtering, Rician noise.

I.INTRODUCTION

The brain tumor is a clot of the cell which can be or non-cancerous. Early detection of tumor is essential to carry out medical treatment. Most of the doctors prefer MRI images of the brain to carry out analysis. These MRI images are highly affected by Rician noise[4]. This noise affected images if used for further processing, it might affect its accuracy. Denoising of an image hence plays a very important role in image processing.

Denoising is carried out byfiltering techniques. There are various filtering techniques available but the resultant filtered image is prone to edge blurring and over smoothing. Maintaining the image features while carrying out image filtering is of extreme importance. Image denoising can be divided into two domain spatial and frequency. In the spatial domain, the operation takes place directly on the pixel while in a frequency domain operation is carried out on Fourier transform of a respective image[10]. The spatial domain denoising techniques include the Mean filter, Median filter, Average filter, Weiner filter while frequency domain denoising includes Fourier transform and Wavelet transform [3].