Abstract:
Embedding of secret information in a digital image will definitely introduce some noise or modulate the
image signal in some way. A good steganographic method ensures that such noise is not perceptible or is
minimal in order to maintain the fidelity of the vessel. The Bit Plane Complexity Segmentation (BPCS)
method uses the Canonical Gray Coded (CGC) bits of the complex bit plane blocks of a vessel image for
embedding secret information. Though this guarantees a high payload capacity, it can potentially
compromise the fidelity of the vessel particularly in its high order bits. To ensure that the vessel is evenly
modulated and hence increase imperceptibility of the embedded data, this paper suggests a tweaking of the
BPCS embedding procedure by employing a random selection of the CGC bits in the noisy regions of the
vessel. Additionally, in order to boost the vessel’s robustness against compression and other image
processing activities, the proposed embedding procedure does not utilize the 0 (zero) CGC bit plane, which
is normally targeted for removal by such activities. Results from the experiments carried out showed that
stego images from the proposed method had improved signal to noise ratios compared to those from the
traditional BPCS method.
