LAUSR

Many noise removal algorithms have been introduced so far for denoising of MRI images like a bilateral filter, wavelet transform, maximum likelihood, non-local means, etc. All the existing methods suffer from some drawbacks or limitations. This research proposed a novel denoising technique for MRI images based on the advanced NLM method with non-subsampled shearlet transform (NSST). NSST is a combination of non-subsampled pyramid filters and non-subsampled shearing filters. NLM filter is modified for better noise detection and calculation of denoised value. The parameters are set to provide maximum output and high-quality denoising. The proposed methods are compared with existing state-of-the-art methods like standard NLM filter, fast NLM, LMMSE, NLML, etc. The analysis is based on values of PSNR, SSIM and RMSE. The results show that the proposed methods remove the noise much effectively as compared to existing denoising methods.