Earticle

초록

영어

A blind steganalysis based on regional correlation and second-order Markov transition probability matrix is proposed for JPEG images. By analyzing the region correlation of JPEG image, the Markov transition probability matrix is used to capture the correlation of intra-block and inter-block DCT coefficients. In addition, the calibrated features are extracted from the calibrated images. The difference features between the original and calibrated images are used for training and classifying. Three different image libraries are used to detect the five kinds of typical JPEG steganography schemes. Experimental results show that, in comparison to the several effective steganalysis, the proposed scheme improves the detection accuracy on some JPEG-based steganography schemes, including outguess, Steghide and MB1.

목차

Abstract
 1. Introduction
 2. Feature Generation
  2.1. Image Scanning Patterns
  2.2. Intra-block Features
  2.3. Inter-block Features
  2.4. Image Calibration
 3. Experimental Results
  3.1. Image Database and JPEG Steganography Tools
  3.2. Experimental Results
 4. Conclusion
 Acknowledgements
 References

키워드

저자

• Su Jinyang [ College of information and engineering, China Jiliang university, Hangzhou, 310018 ]
• Zeng Xianting [ College of information and engineering, China Jiliang university, Hangzhou, 310018 ]
• Wang Lei [ College of information and engineering, China Jiliang university, Hangzhou, 310018 ]

참고문헌

발행기관

간행물

표지보기 관심저널 등록

• 간행물명

  International Journal of Security and Its Applications

• 간기

  격월간

• pISSN

  1738-9976

• 수록기간

  2008~2016

• 등재여부

  SCOPUS

• 십진분류

  KDC 505 DDC 605