Title :
Coding artifacts reduction method based on a perceptually-weighted variational approach
Author :
Do, Q.B. ; Luong, M. ; Beghdadi, A.
Author_Institution :
L2TI, Inst. Galilee, Univ. Paris 13, Villetaneuse, France
Abstract :
A new adaptive post-processing method for reducing blocking and ringing effects is proposed. This approach exploits some characteristics of the human visual system (HVS), namely the just noticeable contrast (JNC) and the luminance adaptation. The method is based on a perceptually-weighted total variation regularization approach constrained by the knowledge of a prior information derived from some structural features maps extracted from the image. The performance of the proposed method is evaluated objectively and subjectively in terms of blocking and ringing visibility. The proposed method is also compared to the state of arts.
Keywords :
brightness; feature extraction; image coding; visual perception; a prior information; adaptive post processing method; blocking effect; coding artifacts reduction method; features maps extraction; human visual system; just noticeable contrast; luminance adaptation; perceptually weighted variational approach; ringing effect; Art; Data mining; Feature extraction; Humans; Smoothing methods; TV; Transform coding; Video compression; Virtual manufacturing; Visual system; Blocking artifact; JPEG; Total Variation; perceptual analysis; post-processing; ringing artifact;
Conference_Titel :
Signal Processing and Information Technology (ISSPIT), 2009 IEEE International Symposium on
Conference_Location :
Ajman
Print_ISBN :
978-1-4244-5949-0
DOI :
10.1109/ISSPIT.2009.5407480