Abstract
In recent years, the Internet has become an indispensable part of life with the highly developed technology, so we cannot ignore the issue about network security. In 2008, Liu proposed a JPEG steganographic algorithm of higher capacity, Complementary Embedding Strategy (CES), and this algorithm can resist several steganographic attacks, such as χ2 attack and S family attacks. And JPEG image format is widely used in a digital camera and in the Internet transmission. In view of this, we choose CES to slightly modify and further increase capacity. In this thesis, we review the concept of steganography and several well-known existing steganographic algorithms first. Then, we present our proposed method, the experimental results and conclusion. Because E1 algorithm of CES may over-modify the selected quantized Discrete Cosine Transform (qDCT) coefficient in the last step to fix the qDCT distribution, we replace the coefficient modification in last step of E1 algorithm of CES with image pre-processing to slightly improve the PSNR value of the stego-image. This section is called modified CES. In addition, if the number of bits of the secret message is larger than the available capacity of modified CES, we will execute a method of raising embedding capacity to increase the available capacity of the cover-image. This method, while it satisfactorily increases the capacity, incurs a slight tradeoff of decreased PSNR value. We test different steganographic algorithms by six 512×512 images with 256 gray levels and show these images’ experimental results of PSNR value and capacity.