A new adaptive noise estimator for PDE-based MR images denoising
Mostafa Heydari, Mohammadreza Karami · 2015
Among different methods of image denoising, PDE (Partial Differential Equation) based denoising attracted much attention in the field of medical image processing. The benefit of PDE-based denoising methods is the ability to remove the noise as well as preserving edge through Anisotropic Diffusion (AD). Although, AD filtering such as Perona-Malik (P-M) model is widely used for MR Image enhancement, but this filtering is nonoptimal for MR Images that have Rician noise. Thus, this filter should be fitted with Rician noise. One of the most useful AD models that are fitted with Rician noise is AADM (automatic parameter selection anisotropic diffusion for MR Images). In this paper, we propose a new adaptive method to estimate standard deviation of noise and correct the bias error. It causes that the performance of AADM model improves. Experimental results show that when we apply our proposed estimator to AADM method, its performance (such as SNR and edge-preserving) to remove Rician noise in MR Images improves, effectively.