%0 Conference Proceedings
%4 sid.inpe.br/sibgrapi/2012/07.15.22.08
%2 sid.inpe.br/sibgrapi/2012/07.15.22.08.42
%@doi 10.1109/SIBGRAPI.2012.12
%T Contextual filtering of CT images using Markovian Wiener filters with a Non Local Means approach for statistical estimation
%D 2012
%A Salvadeo, Denis Henrique Pinheiro,
%A Mascarenhas, Nelson Delfino d'Ávila,
%A Levada, Alexandre Luís Magalhães,
%@affiliation Federal University of São Carlos (UFSCar)
%@affiliation Federal University of São Carlos (UFSCar)
%@affiliation Federal University of São Carlos (UFSCar)
%E Freitas, Carla Maria Dal Sasso ,
%E Sarkar, Sudeep ,
%E Scopigno, Roberto ,
%E Silva, Luciano,
%B Conference on Graphics, Patterns and Images, 25 (SIBGRAPI)
%C Ouro Preto, MG, Brazil
%8 22-25 Aug. 2012
%I IEEE Computer Society
%J Los Alamitos
%S Proceedings
%K Wiener filter, Fisher Information, Markov Random Fields, Non Local Means approach, image denoising, Computed Tomography.
%X Recently, investigations on medical imaging have been indicating a strong correlation between cases of cancers and the increasing number of Computed Tomography (CT) exams, mainly due to high radiation doses to which patients are exposed during the data acquisition process. Thus, there is a need to reduce the radiation doses whereas still keeping satisfactory quality images for diagnosis. In this paper, we propose to filter noise in CT images using contextual versions of Wiener Filter such as Generalized Wiener Filter (GWF) and Non Local Means approach for parameter estimation. Experiments show that the proposed methods are promising, since they provide good results with no significant increase in the computational cost.
%@language en
%3 101966.pdf