Software and hardware exploition of volterra and optimized volterra filter for denoising mri images

Abstract

Digital images play a pivotal rolein our daily life and are broadly newlineused in many applications resembling medical applications, remote sensing, newlineastronomy, satellite communication, etc. In digital image processing, denoising newlineis an important process more important than any other tasks like newlineimage enhancement and image restoration. Image de-noising is the most newlineeffective process for achieving both noise reduction and feature extraction. newlineThe function of de-noising is to remove the noise while maintaining the edges newlineand other detailed features as far as possible. In the image, the noise gets newlinegenerated during image acquisition, transmission, reception, storage and newlinerepossession processes. The produced noise will degrade the information and newlineinsights the quality of an image. Importantly, in medical applications, the newlineMRI image owns a critical part to identify the correct diagnosis and types of newlinediseases. newlineBut the noise in MRI images leads to the fault diagnosis. Hence, noise newlineremoval is the major task in medical images. To recover the original image newlinevarious noise removal techniques were suggested such as the Linear newlineMinimum Mean Squared Error method (LMMSE), histogram-based denoising newlineand wiener filter. The noises considered in this thesis are Gaussian noise, newlineRandom field noise, and Rician noise. Some spatial filtering and transformdomain newlineimage filtering algorithms have been listed to curb multiple noises newlinefrom the MRI image.In literature, many proficient image filters are initiated newlinethat achieve well. Even though they remove the noise from the image as newlinemuch, the output image will be blurred and causes spatial filtering newline