An Edge Preserving Denoising.Technique for MR Images using Curvelet Transform

摘要

This paper presents a curvelet based approach for the denoising of magnetic resonance (MR) images. Curvelet transform is a new multi-scale representation suited for objects which are smoothened away from discontinuities across curves. It was developed by Candes and Donoho. These digital transforms are applied to the denoising of standard MR images embedded in Gaussian noise, random noise and Poisson noise, in the tests reported here, simple thresholding of the Curvelet coefficients is very competitive with state-of-the-art techniques based on wavelet transform methods. Moreover, the curvelet reconstructions exhibit higher perceptual quality than wavelet-based reconstructions, offering visually sharper images and, in particular, higher quality recovery of edges and of faint linear and curvilinear features. Since medical images have several objects and curved shapes, it is expected that curvelet transform would be better in their denoising. The simulation resultsshow that the proposed curvelet method outperforms the wavelet method in the denoising of both MR images in visual quality and the peak signal to noise ratio (PSNR) points of view.