Design of Reversible JPEG Image Compression Scheme for Data Hiding

This paper presents a new Steganography technique, In this paper, we propose a new high-capacity reversible data hiding method for JPEG-compressed images. This method is based on modifying the quantization table and quantized discrete cosine transformation (DCT) coefficients. Some elements of the quantization table are divided by an integer while the corresponding quantized DCT coefficients are multiplied by the same integer and added by an adjustment value to make space for embedding the data. By analyzing the effect of each single quantized DCT coefficient on the image quality, an embedding sequence is chosen to help control the increase of file size after hiding the data meanwhile the PSNR value between the original uncompressed image and stego JPEG image is high. Experimental results show that the proposed method achieves both high capacity and high image quality. The simulations are taken for the MATLAB standard images of Lena, baboon, Barbara, Jet, peppers, with images size of 512x512 size, and the data hidden is of 1 Kb. The PSNR observe between 36 to 38 for those images and that is significantly better than base works.