JPEG Image Authentication through Reversible Data Hiding using Differential Histogram of Quantized Coefficients

Since digital contents are easy to copy and manipulate without the loss of quality, many social issues related to forgery and illegal distribution are raised to protect the copyright of contents. Reversible data hiding is an effective way which keeps the quality of the contents and embeds copyright information into contents itself. Although compression is inevitable for multimedia, few reversible data hiding researches against compression have been performed. In this paper, we propose a novel reversible data hiding algorithm that is applied to authenticate commonly used JPEG images. The differential histogram of the quantized coefficients among adjacent blocks is calculated and the 64 bits authentication code is embedded by shifting this differential histogram during JPEG compression. The embedded code is extracted by reversing the shifting process and applied to check the integrity. Without any quality degradation, the original JPEG image can be reconstructed by calculating original quantized coefficients. The proposed algorithm was tested under various compression rates. The PSNR and compression ratio of the watermarked JPEG images were 32.33dB and 89.69% whose differences with the original JPEG images were 3.24dB and 2.58%.