3D Adaptive SMCM Filtering of CT Images

In this paper a new implementation of the Saint-Marc-Chen-Medioni (SMCM) filter is proposed in a 3D variant. It uses adaptive selection of the threshold
$k$
, found from a 2D projection of the input image. The algorithm has been tested over Computed Tomography (CT) images with Additive White Gaussian Noise (AWGN) at variance levels 0.01 and 0.001. It leads to better quality of filtered images in terms of Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM), compared to 3D Gaussian and 3D Averaging filters. The proposed 3D SMCM filter is considered applicable for filtering of other types of 3D images, such as Magnetic Resonance Images (MRI).