Prof. Ing. Francescomaria Marino

Associate Professor

DEE Politecnico di Bari

JASTEG2000 is a visual environment, developed by Domenico Introna and Francescomaria Marino whose name is two-fold:

  1. It is a tribute to the well known freeware JASPER [1], used for developing the codec;
  2. it recalls its main goal: “STEGanography for JPEG2000 cover images”.

In fact, JASPEG2000 provides a powerful codec with the possibility of steganography in JPEG2000 cover images.

The word steganography comes from two Greek words: στεγανοσ (i.e., steganos=hidden) and γραφια (i.e., graphia=writing). The join of these words describes the concept of a communication which hides not only its contents, but also its existence. This is well different from cryptography, which makes incomprehensible the meaning of a communication, though this one can be seen by everybody.

This concept has already been exploited in the past. Probably, the eldest example of steganography is narrated by the Greek historian Herodotos (486-425 B.C.), who writes that a noble Persian, Histianeus, shaved the hairs of a slave and tattooed on his glabrous head a secret message. The messenger left only after his hairs were grown again. By this way, none could suspect the presence of a message, that was so hidden by the hairs. When the messenger reached the addressees, his hairs were once again shaved and the message could be normally read.

A more recent example of steganography are the photographic micro-dots, which were defined by the FBI Director, J. E. Hoover as “the enemy’s masterpiece of spionage”. During the Second World War, the Germans used very high quality micro photos as media for a big amount of data. These micro photos could be reduced to the size of a little “dot” and hidden in unsuspicious type-written letters, for instance, as dots of several “i”. In order to read the message, the addressees had only to magnify the micro photos.

The latter example provides the essence of many steganographic techniques, which usually exploit a second perceptible message (“cover”), having disjoined meaning by the secret message. The cover works as a “Trojan horse”, being a container of the secret message [2-4]. In this context, digital images are excellent “covers”, since the bits of a secret message can be superimposed, as slight noise, to the bits coding the cover image.

At the best of our knowledge, only two steganographic algorithms, dealing with the JPEG2000 image standard coding [5-8], have been proposed.

These two state-of-the-art methods are the algorithm proposed by Pochi-Su and C.-C. Jay Kuo [9] and JPEG2000-BPCS (Bit-plane Complexity Segmentation) proposed by H. Noda et al. [10]. These two methods follow different philosophies. Briefly, method [9] has the goal of keeping constant the size of the cover image file, pre and post the embedding. As a consequence, it is necessary bounded for what concerns the embedding capacity. Conversely, JPEG2000-BPCS aims at getting high capacity, even with a low distortion, though it gives less importance to the eventual post-embedding growth.

JASTEG2000 implements a High Embedding/Low Distortion (HELD) steganographic method described in [11], which comes in four slightly different versions (HELD.1, HELD.2, HELD.3 and HELD.4), each one matching different compromises between post-embedding growth and distortion. In the worst cases, HELD gets up to 35%-45% embedding rate with approximately 2 dB of distortion, and 30%-40% embedding rate with less than 4 dB of distortion, when applied on classical benchmarks codedat 0.5 bpp and 1.0 bpp, respectively. These results outperform those achieved by methods [9] and [10].

Alternatively, it offers the possibility of performing steganography without growth of the cover image file (method Growth 0), thanks to Dr. Pochyi-Su (National Central University, Taiwan), which has kindly provided the source code of the algorithm described in [9], with the permission of embedding it into JASTEG2000.

JASTEG2000 may be requested for research purposes writing to marino@poliba.it.

Figures

JPEG2000 Console.

pop-up setting panel for JPEG2000 coding.

Growth Rate vs Embedded Message for Mandrill, Barbara and Lena coded at 0.5 and 1.0 bpp: HELD.1

Growth Rate vs Embedded Message for Mandrill, Barbara and Lena coded at 0.5 and 1.0 bpp: HELD.2

Growth Rate vs Embedded Message for Mandrill, Barbara and Lena coded at 0.5 and 1.0 bpp: HELD.3

Growth Rate vs Embedded Message for Mandrill, Barbara and Lena coded at 0.5 and 1.0 bpp: HELD.4

PSNR vs Embedding Rate for Mandrill, Barbara and Lena coded at 0.5 and 1.0 bpp: HELD.1

PSNR vs Embedding Rate for Mandrill, Barbara and Lena coded at 0.5 and 1.0 bpp: HELD.1

PSNR vs Embedding Rate for Mandrill, Barbara and Lena coded at 0.5 and 1.0 bpp: HELD.2

PSNR vs Embedding Rate for Mandrill, Barbara and Lena coded at 0.5 and 1.0 bpp: HELD.3

PSNR vs Embedding Rate for Mandrill, Barbara and Lena coded at 0.5 and 1.0 bpp: HELD.4

Lena, coded at 1.0 bpp, before the embedding.

Lena, after the embedding (Embedding rate=30%).

Barbara, coded at 1.0 bpp, before the embedding.

Barbara, after the embedding (Embedding rate=30%).

Baboon, coded at 1.0 bpp, before the embedding.

Baboon, after the embedding (Embedding rate=30%).

Growth Rate vs Embedded Message for Barbara coded at 0.5 and 1.0 bpp. A grey triangle and a grey square denote the values obtained by JPEG-2000 BPCS, as reported in [10].

PSNR vs Embedding Rate in case of steganography by HELD and by Algorithm [9], for Lena and Barbara coded at 1.0 bpp.

PSNR vs Embedding Rate in case of JPEG2000-BPCS (as shown in Fig. 2 of [10]).

PSNR vs Embedding Rate in case of JPEG2000-BPCS (as shown in Fig. 2 of [10]).

PSNR vs Embedding Rate in case of HELD.4.

References

[1]   http://www.ece.uvic.ca/~mdadams/jasper/
[2]    W. Bender, D. Gruhl, N. Morimoto and A. Lu., “Techniques for Data Hiding,” I.B.M. Systems Journal, vol.35, no.3-4, 1996, pp. 313-336.
[3]   F.A.P. Petitcolas, R.J. Anderson and M.G. Kuhn, “Information Hiding – A Survey,” Proceedings of the IEEE, vol.87, no.7, 1999, pp. 1062-1078.
[4]   S. Katzenbeisser and F. A. P. Petitcolas, Information Hiding Techniques for Steganography and Digital Watermarking. Norwood, MA: Artech House, 2000.
[5]   C. Christopoulos, A. Skodras and T. Ebrahimi, “The JPEG2000 still image coding system: An overview,” IEEE Trans. on Consumer Electronics, vol. 46, no. 4, Nov. 2000, pp. 1103-1127.
[6]   http://www.jpeg.org/JPEG2000.html
[7]   M. D. Adams, “The JPEG-2000 still image compression standard,” ISO/IEC JTC 1/SC 29/WG 1, Tech. Rep., Sep. 2001.
[8]   M. Rabbani and R. Joshi, “An overview of JPEG 2000 still image compression standard,” Signal Process.: Image Commun., 2002, vol. 17, pp. 3–48.
[9]   Po-Chyi Su and C.-C.J. Kuo, “Steganography in JPEG2000 compressed images,” IEEE Transactions on Consumer Electronics, V. 49 , N. 4, Nov. 2003, p. 824-832.
[10]   H. Noda, J. Spaulding, M.N. Shirazi and E. Kawaguchi, “Application of bit-plane decomposition steganography to JPEG2000 encoded images,” IEEE Signal Processing Letters, V. 9, N. 12, Dec. 2002, pp. 410-413.
[11]   D. Introna and F. Marino, “HELD: A High Embedding/Low Distortion Steganographic Algorithm for JPEG2000 Coded Images,” submitted to IET Information Security.

poliba Site Map | mova::webdesign | ©2006 Politecnico di Bari