An efficient example-based method for CT image denoising based on frequency decomposition and sparse representation

In this paper we present an efficient example-based method for Gaussian denoising of CT images. In the proposed method, an image is considered as a sum of the three frequency bands: low-band, middle-band and high-band. We assume that the noise component is often mixed into the middle-band and the high-band in order to better preserve the high-frequency details in the image we perform denoising on these two bands. The method is based on a sparse representation model in which a set of standard images is used to construct the example dictionaries. The experimental results demonstrate that the proposed denoising method can preserve well the high-frequency details. The objective and subjective comparisons also show that the proposed our method outperforms other state-of-the-art denoising methods.