• DocumentCode
    1576801
  • Title

    A multi-scale image inpainting algorithm based on GMRF model

  • Author

    Wang, Rui ; Gu, Dongbing ; Liu, Guangwen ; Sun, Junxi

  • Author_Institution
    Sch. of Electron. & Inf. Eng., Changchun Univ. of Sci. & Technol., Changchun, China
  • fYear
    2009
  • Firstpage
    1844
  • Lastpage
    1848
  • Abstract
    A novel approach to image inpainting is introduced in this paper. The novelty lies in the combination of the MRF inpainting technique and a multiscale information fusion mechanism. This combination uses both local information via the MRF technique and global information via the multi-scale mechanism to improve the image inpainting quality. The proposed algorithm consists of three steps. In the first step, the image to be inpainted is decomposed into multiple levels via a multi-scale decomposition. In the second step, the image of each level is inpainted based on the MRF model. Finally, the multi-scale inpainting results of all levels with different scales are fused. The experimental results show that the proposal inpainting algorithm achieves fairly good results.
  • Keywords
    Gaussian processes; Markov processes; image fusion; GMRF model; Gaussian Markov random field model; image inpainting quality; multiscale image decomposition; multiscale image inpainting algorithm; multiscale information fusion mechanism; Biomimetics; Computer graphics; Filling; Image restoration; Markov random fields; Pixel; Probability density function; Proposals; Robots; Sun;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Robotics and Biomimetics (ROBIO), 2009 IEEE International Conference on
  • Conference_Location
    Guilin
  • Print_ISBN
    978-1-4244-4774-9
  • Electronic_ISBN
    978-1-4244-4775-6
  • Type

    conf

  • DOI
    10.1109/ROBIO.2009.5420444
  • Filename
    5420444