Laser & Optoelectronics Progress, Volume. 56, Issue 2, 021005(2019)

Criminisi Image Inpainting Algorithm Based on Rough Data-Deduction

Ning Zhou* and Zhaozhao Zhu
Author Affiliations
  • School of Electronics and Information Engineering, Lanzhou Jiaotong University, Lanzhou, Gansu 730070, China
  • show less
    Figures & Tables(9)
    Schematic of improved matching block search
    Schematic of rough data-deduction of image
    Flow chart of improved Criminisi algorithm
    Rectangular inpainting effects. (a) Original image; (b) damaged image; (c) repair result by Criminsi algorithm; (d) idea repair result by rough data-deduction
    Inpainting effects of image removal. (a) Original image; (b) damaged image; (c) repair result by Criminsi algorithm; (d) idea repair result by rough data-deduction
    Inpainting effects of color image. (a) Original image; (b) damaged image; (c) repair result by Criminsi algorithm; (d) idea repair result by rough data-deduction
    Inpainting effects of animal removal. (a) Original image; (b) damaged image; (c) repair result by Criminsi algorithm; (d) idea repair result by rough data-deduction
    Peak signal-to-noise ratio comparison after image inpainting
    • Table 1. Comparison of physical memory usage after image restoration

      View table

      Table 1. Comparison of physical memory usage after image restoration

      Figure numberPhysical memory occupied byCriminisi algorithm /MbPhysical memory occupiedby rough data-deduction/Mb
      Fig.4(b)68
      Fig.5(b)1019
      Fig.6(b)1017
      Fig.7(b)3133
    Tools

    Get Citation

    Copy Citation Text

    Ning Zhou, Zhaozhao Zhu. Criminisi Image Inpainting Algorithm Based on Rough Data-Deduction[J]. Laser & Optoelectronics Progress, 2019, 56(2): 021005

    Download Citation

    EndNote(RIS)BibTexPlain Text
    Save article for my favorites
    Paper Information

    Category: Image Processing

    Received: Jul. 5, 2018

    Accepted: Aug. 3, 2018

    Published Online: Aug. 1, 2019

    The Author Email: Zhou Ning (zhouning@mail.lzjtu.cn)

    DOI:10.3788/LOP56.021005

    Topics