• DocumentCode
    3543575
  • Title

    An iterative least-square training method for classification-based motion adaptive temporal filtering

  • Author

    Kim, Sung Deuk ; Lim, Kyoung Won

  • Author_Institution
    Andong Nat. Univ., Andong, South Korea
  • fYear
    2010
  • fDate
    9-13 Jan. 2010
  • Firstpage
    389
  • Lastpage
    390
  • Abstract
    This paper presents a motion-adaptive temporal filtering scheme based on trained least-square filtering. A novel iterative method is utilized in the training stage to find the optimal motion-adaptive temporal filter coefficient. The proposed approach shows optimal noise reduction performance in the LMS sense without introducing any blur.
  • Keywords
    IIR filters; adaptive filters; image classification; image denoising; image motion analysis; iterative methods; least mean squares methods; optimisation; IIR filter; classification-based motion adaptive temporal filtering scheme; iterative least-square training method; least square optimization method; least-square filtering; optimal motion-adaptive temporal filter coefficient; optimal noise reduction performance; Adaptive filters; Detectors; Error correction; Filtering; Iterative methods; Least squares methods; Motion detection; Noise reduction; Optimization methods; Piecewise linear techniques;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Consumer Electronics (ICCE), 2010 Digest of Technical Papers International Conference on
  • Conference_Location
    Las Vegas, NV
  • Print_ISBN
    978-1-4244-4314-7
  • Electronic_ISBN
    978-1-4244-4316-1
  • Type

    conf

  • DOI
    10.1109/ICCE.2010.5418818
  • Filename
    5418818