• DocumentCode
    2760008
  • Title

    The Correction of Intensity Flicker in Archived Film

  • Author

    Zhang, Ranran ; Wu, Jiong ; Ding, Youdong ; Hao, Pengyi

  • Author_Institution
    Dept. of Comput. Sci. & Technol., Shanghai Univ., Shanghai, China
  • Volume
    2
  • fYear
    2009
  • fDate
    25-26 July 2009
  • Firstpage
    402
  • Lastpage
    405
  • Abstract
    The detection and restoration of intensity flicker is a common artifact in the restoration of archived film. Considering the interrelation between the successive frames in the image sequences, a novel approach based on motion compensation is proposed. Three steps search algorithm is introduced to estimate motion vectors, which can detect motion regions more accurately. Besides, threshold method is used to detect flickers, avoid non-flicker regions being corrected, and the efficiency has been improved with shortened correction time. The experimental results show that with the accurate correction of intensity flicker the proposed algorithm can preserve the wanted effects with similar features, and shorten the time of intensity flicker correction.
  • Keywords
    entertainment; image segmentation; image sequences; motion compensation; motion estimation; search problems; archived film; image sequence; intensity flicker correction; intensity flicker detection; intensity flicker restoration; motion compensation; motion region detection; motion vector estimation; search algorithm; threshold method; 1f noise; Additive noise; Computer science; Equations; Gaussian noise; Image restoration; Motion detection; Motion estimation; Motion pictures; Parameter estimation; flicker correction; intensity flicker; motion estimation; three steps search algorithm;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Information Technology and Computer Science, 2009. ITCS 2009. International Conference on
  • Conference_Location
    Kiev
  • Print_ISBN
    978-0-7695-3688-0
  • Type

    conf

  • DOI
    10.1109/ITCS.2009.220
  • Filename
    5190264