• DocumentCode
    2456066
  • Title

    Flow-like textures with line-like structures images inpainting by improved tensor diffusion model

  • Author

    Cui Xuehong

  • Author_Institution
    Coll. of Inf. Sci. & Technol., Qingdao Univ. of Sci. & Technol., Qingdao, China
  • fYear
    2010
  • fDate
    24-27 Aug. 2010
  • Firstpage
    1773
  • Lastpage
    1777
  • Abstract
    For inpainting of the flow-like textures with line-like structures images, we present an model based on tensor diffusion. The proposed model is to reconstruct a damaged original image by acting along the texture´s direction and the vertical texture´s direction which is determined by the texture´s trend of the local image structure (This local image structure is measured by the so-called structure tensor). In order to guarantee better results, the intensity of the diffusion along the texture´s direction should be much greater than that of the diffusion along the vertical texture´s direction. The nonnegativity discretization method and optimized rotation invariance method is used to calculate the proposed model. It was shown in experiments that this model can get good results for inpainting the flow-like textures with line-like structures images that contain nicks or small broken areas.
  • Keywords
    image reconstruction; image texture; tensors; flow-like texture; image reconstruction; line-like structure image inpainting; nonnegativity discretization method; optimized rotation invariance method; tensor diffusion model; vertical texture direction; Anisotropic magnetoresistance; Coherence; Eigenvalues and eigenfunctions; Equations; Image reconstruction; Mathematical model; Tensile stress; flow-like texture; image inpainting; line-like structures; structure tensor;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Computer Science and Education (ICCSE), 2010 5th International Conference on
  • Conference_Location
    Hefei
  • Print_ISBN
    978-1-4244-6002-1
  • Type

    conf

  • DOI
    10.1109/ICCSE.2010.5593835
  • Filename
    5593835