TitleHeNLM-3D: A 3D Computer Tomography Image Denoising Algorithm
Publication typeConference paper
Author(s)N. Mamaev, A. Lukin, D. Yurin
Publication dateMay 2014
ConferenceThe 12th International Conference on Pattern Recognition and Information Processing
Address28–30 May 2014, Minsk, Belarus
PublisherUnited Institute of Informatics Problems of the National Academy of Sciences of Belarus
AbstractA 3D filtering algorithm for CT images is proposed (HeNLM-3D). It is an extension of the known Local Jets method (LJNLM-LR) and our previous 2D algorithm (HeNLM). The proposed algorithm is based on the expansion of pixel neighborhoods into Hermite functions which form the orthonormal system. Tests on real CT images have shown that HeNLM-3D performs filtering better in high-detailed areas in comparison with LJNLM-LR and classical NLM method.

Show BiBTeX entry