Data hiding method in a joint photographic experts group (JPEG) image compression

We propose a data hiding in a jpeg image compression algorithm which combines a image compression followed by a data hiding. The main features of a proposed method includes the use of DCT, sub-block arrangement, quantization, run lengh encode. At first discrete cosine transform applied to the cover image after that data is embedded into the encrypted image in the location of ac and dc component of a truple. truple is a 3X3 block of a 8X8 compressed image sub block. Experimental result demonstrate the performance of MATLAB standard images for QF=50, 60….90. This scheme brings three benefits (i) lower mean square error; (ii) higher PSNR for QF=50 and 60 (iii) higher embedding rate. The paper justifies the proposed algorithm achieves significantly higher embedding rate and PSNR..