• DocumentCode
    981396
  • Title

    Strong-continuation, contrast-invariant inpainting with a third-order optimal PDE

  • Author

    Bertalmio, Marcelo

  • Author_Institution
    Dept. de Tecnologia, Univ. Pompeu Fabra, Barcelona, Spain
  • Volume
    15
  • Issue
    7
  • fYear
    2006
  • fDate
    7/1/2006 12:00:00 AM
  • Firstpage
    1934
  • Lastpage
    1938
  • Abstract
    PDE-based image inpainting has become a very active area of research after the pioneering works of Masnou and Morel, Bertalmı´o et al., and Ballester et al. In this paper, we take a different approach, inspired by the excellent work of Caselles et al. We view the inpainting problem as a particular case of image interpolation in which we intend to propagate level lines. Expressing this in terms of local neighborhoods and using a Taylor expansion we derive a third-order PDE that performs inpainting. This PDE is optimal in the sense that it is the most accurate third-order PDE which can ensure continuation of level lines. The continuation is strong, allowing the restoration of thin structures occluded by a wide gap. The result is also contrast invariant. This is a novel PDE, which, in both its accuracy and contrast invariance, outperforms the approaches cited above.
  • Keywords
    image restoration; interpolation; partial differential equations; PDE-based image inpainting; Taylor expansion; contrast invariance; image interpolation; local neighborhoods; partial differential equations; strong-continuation contrast-invariant inpainting; thin structure restoration; third-order optimal PDE; Art; Geometry; Helium; Image processing; Image restoration; Image segmentation; Interpolation; Layout; Partial differential equations; Taylor series; Image restoration; inpainting; partial differential equations (PDEs); Algorithms; Artifacts; Computer Graphics; Computer Simulation; Image Enhancement; Image Interpretation, Computer-Assisted; Information Storage and Retrieval; Models, Statistical; Numerical Analysis, Computer-Assisted; Paintings; Signal Processing, Computer-Assisted; Stochastic Processes;
  • fLanguage
    English
  • Journal_Title
    Image Processing, IEEE Transactions on
  • Publisher
    ieee
  • ISSN
    1057-7149
  • Type

    jour

  • DOI
    10.1109/TIP.2006.877067
  • Filename
    1643700