Cryptography Reference
In-Depth Information
Conclusions from the Experiments
Throughout varieties of embedding and extracting experiment using BMP,
PNG, and JPEG, we confirmed the followings.
(A) BPCS-embedding method has a very large embedding capacity. Espe-
cially, in the case of PNG image, the capacity turned out to be 65% in
most cases. When the embedded data are all text files, because the LHA
compression program can reduce a text file into 1/3 of its original size, it
is possible to embed almost 200% larger text data into one PNG image.
(B) Embedding capacity according to our JPEG embedding program was
found to be 10% on average.
(C) Image quality after embedding did not change visually even in the case
when we embedded the maximum capacity in both BPCS and JPEG
method.
(D) Execution time of BPCS embedding and extracting were quite fast. Of
course it depends on the size of the vessel image and the size of the
embedding data. It was also true for JPEG embedding and extracting.
(E) The embedded data are very fragile in both BPCS and JPEG method.
Changing color data by 1 bit on one single pixel of a stego image always
hampered the extracting operation.
References
1. Hall, E.L. (1979): Computer image processing and recognition. Academic Press,
New York
2. Jain, A.K. (1989): Fundamentals of digital image processing. Prentice Hall, En-
glewood Cliffs, NJ
3. Kawaguchi, E., Endo, T., Matsunaga, J. (1983): Depth-first picture expression
viewed from digital picture processing. IEEE Trans. on PAMI
5
373384
4. Kawaguchi, E., Taniguchi, R.: Complexity of binary pictures and image thresh-
olding An application of DF-Expression to the thresholding problem. Pro-
ceedings of 8th ICPR,
2
, 12211225
5. Kawaguchi, E., Taniguchi, R. (1989): The DF-Expression as an image thresh-
olding strategy. IEEE Trans. on SMC,
19
, 13211328
6. Kamata, S., Eason, R.O., Kawaguchi, E. (1995): Depth-first Coding for multi-
valued pictures using bit-plane decomposition. IEEE Trans. on Commun.,
43
,
19611969
7. Kawaguchi, E., Niimi, M. (1988): Modeling digital image into informative and
noise-like regions by complexity measure. Information Modelling and Knowledge
Bases XV, IOS Press, 255265
8. Nozaki, K., Noda, H., Kawaguchi, E., Eason, R. (2005): A model of unforge-
able digital certificate document system. Information Modelling and Knowledge
Bases XVI, IOS Press, 203210
9. Niimi, M., Noda, H., Kawaguchi, E. (1999): Steganography based on region
segmentation with a complexity measure. Systems and Computers in Japan,
30
, 19
