A Steganography Scheme on JPEG Compressed
Cover Image
with High Embedding Capacity
Arup Kumar Pal1, Kshiramani Naik1, and Rohit Agarwal2
1Department of Computer Science and Engineering, Indian
Institute of Technology(ISM), India
2Department
of Computer Science and Engineering, JSS Academy of
Technical Education, India
Abstract: Joint
Photographic Experts Group (JPEG) is one of the widely used lossy image
compression standard and in general JPEG based compressed version images are
commonly used during transmission over the public channel like the Internet. In
this paper, the authors have proposed a steganography scheme where the secret
message is considered for embedding into the JPEG version of a cover image. The
steganography scheme initially employs block based Discrete Cosine
Transformation (DCT) followed by some suitable quantization process on the
cover image to produce the transformed coefficients. The obtained coefficients
are considered for embedding the secret message bits. In general, most of the
earlier works hide one bit message into each selected coefficient, where hiding
is carried out either directly modifying the coefficients, like employing the
LSB method or indirectly modifying the magnitude of the coefficients, like
flipping the sign bit of the coefficients. In the proposed scheme, instead of
embedding the secret message bits directly into the coefficients, a suitable
indirect approach is adopted to hide two bits of the secret message into some
selected DCT coefficients. As per the conventional approach, the modified
coefficients are further compressed by entropy encoding. The scheme has been
tested on several standard gray scale images and the obtained experimental
results show the comparative performance with some existing related works.
Keywords: Chi-square
attack; (DCT); Histogram; (JPEG); statistical steganalysis; steganography.