• DocumentCode
    1848450
  • Title

    A refining localization watermarking for image tamper detection and recovery

  • Author

    Baotian Cheng ; Rongrong Ni ; Yao Zhao

  • Author_Institution
    Inst. of Inf. Sci., Beijing Jiaotong Univ., Beijing, China
  • Volume
    2
  • fYear
    2012
  • fDate
    21-25 Oct. 2012
  • Firstpage
    984
  • Lastpage
    988
  • Abstract
    In this paper, we propose an efficient watermarking algorithm for image tamper localization and recovery. To enhance the accuracy of tamper localization, the algorithm contains basic authentication and refining authentication. Basic authentication can locate all blocks that contain the tampered pixels by the basic authentication watermark. Refining authentication can exclude some genuine pixels from the tampered block by the refining authentication watermark that is generated based on image disordering. Finally, we fuse the two authentication results and take some post-processing measures to obtain the accuracy authentication result. The recovery watermark is the feature of every block, which is generated using the important DCT coefficients. Theoretical analysis and simulation results show that the proposed algorithm can improve the accuracy of tamper localization. At the same time, the algorithm can also approximately recover the tampered pixels.
  • Keywords
    discrete cosine transforms; image watermarking; object detection; DCT coefficients; efficient watermarking algorithm; genuine pixels; image disordering; image tamper detection; image tamper localization; image tamper recovery; post-processing measures; refining authentication watermark; refining localization watermarking; tamper localization; tampered pixels; theoretical analysis; theoretical simulation; Basic authentication watermark; Image disordering; Post-processing; Refining authentication watermark;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Signal Processing (ICSP), 2012 IEEE 11th International Conference on
  • Conference_Location
    Beijing
  • ISSN
    2164-5221
  • Print_ISBN
    978-1-4673-2196-9
  • Type

    conf

  • DOI
    10.1109/ICoSP.2012.6491744
  • Filename
    6491744