• DocumentCode
    703535
  • Title

    A distributed adaptive block matching algorithm: DIS-ABMA

  • Author

    Vermaut, F. ; Deville, Y. ; Macq, B. ; Marichal, X.

  • Author_Institution
    Dept. of Comput. Sci. & Eng., Univ. Catholique de Louvain, Louvain-la-Neuve, Belgium
  • fYear
    1998
  • fDate
    8-11 Sept. 1998
  • Firstpage
    1
  • Lastpage
    4
  • Abstract
    Variable block size algorithms for motion estimation, like the Adaptive Block Matching Algorithm (ABMA) [10], have been proposed to better match to "objects in motion" compared to the classical BMA [4, 2]. However, the variable block size grid derivation and the related motion estimation relies on a regularization process which implies heavy iterative and inter-dependent computation. Though the paral-lelization of the BMA is si gh forward, the ABMA needs a deeper analysis before its implementation in a distributed environment : this is the goal of this paper. We first designed a modelization of the motion detection of ABMA. This model can lead to several different distributed versions. A specific distributed model, with one master and several slaves, is then described. An implementation of this model has been realized (in C++, using the PVM platform). Experimentations show a linear speedup with respect to the number of processors.
  • Keywords
    image matching; motion estimation; DIS-ABMA; PVM platform; distributed adaptive block matching algorithm; distributed environmental model; interdependent computation; iterative computation; motion detection; motion estimation; regularization process; variable block size algorithm; variable block size grid derivation; Complexity theory; Computational modeling; Data structures; Motion detection; Motion estimation; Prediction algorithms; Video compression; Image processing; distributed algorithms; motion estimation; video coding;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Signal Processing Conference (EUSIPCO 1998), 9th European
  • Conference_Location
    Rhodes
  • Print_ISBN
    978-960-7620-06-4
  • Type

    conf

  • Filename
    7090006