• DocumentCode
    588839
  • Title

    Steganalysis on JPEG Decompressed Bitmaps Revisited

  • Author

    Weiquan Cao ; Jiesi Han

  • Author_Institution
    Southwest Electron. & Telecommun. Technol. Res. Inst. Chengdu, Chengdu, China
  • fYear
    2012
  • fDate
    2-4 Nov. 2012
  • Firstpage
    878
  • Lastpage
    881
  • Abstract
    It´s well known that the spatial ±1 steganography on JPEG decompressed images is relatively easier to detect than those never compressed ones. Though some targeted methods that are special for JPEG decompressed carriers are designed, there are still some drawbacks among them. First, most methods are not perfect or even don´t work for embedding rate as low as 0.0003 bits per pixel, or even lower. Second, few methods are able to estimate embedding rate for LSB matching. Finally, many methods require estimation of the JPEG quality factor, which makes the methods slow and not so robust for big JPEG quality factors. This paper revisited a color based method raised by Westfeld ten years ago, unearthed the reason why it´s so reliable and also extended it to a practical non-parameter quantitative steganalyzer for JPEG decompressed carriers. Error rates of our new method was evaluated in this paper.
  • Keywords
    image coding; image colour analysis; image matching; steganography; JPEG decompressed bitmaps; JPEG decompressed carriers; JPEG decompressed images; JPEG quality factor estimation; LSB matching; color-based method; embedding rate estimation; nonparameter quantitative steganalyzer; spatial steganography; Detectors; Estimation; Image coding; Image color analysis; Q factor; Transform coding; Transforms; LSB matching; decompressed image; information hiding; quantitative steganalysis;
  • fLanguage
    English
  • Publisher
    ieee
  • Conference_Titel
    Multimedia Information Networking and Security (MINES), 2012 Fourth International Conference on
  • Conference_Location
    Nanjing
  • Print_ISBN
    978-1-4673-3093-0
  • Type

    conf

  • DOI
    10.1109/MINES.2012.205
  • Filename
    6405830