This paper proposes a novel data hiding framework based on image hashing
strategy. First, the best-matching blocks for each cover image block are searched in the
existing database, which can be generated by the codebook of vector quantization technique.
Then, a dictionary consisting of different versions for each cover block is constructed by
a set of content-preserving manipulations, and each version is assigned with a unique
binary representation. According to the secret data hidden for each block, corresponding
versions of the block in the dictionary are chosen to substitute the cover block and form the
stego image. Due to the properties of robustness and anti-collision of image hashing, the
secret data can be extracted correctly by comparing the stego block with the blocks in the
reconstructed dictionary. The experimental results show the effectiveness and
exibility
of the proposed framework.
Relation:
International Journal of Innovative Computing, Information and Control, V.9 N.2 : 599-610.
