Error Correction Coding And Optimization Algorithms Applied Research In Image Steganography

With the rapid development and application of Internet, the issue of information security has become visible. Technology of information hiding as a main method of hidden communication and protection of knowledge copyright has been widely used and developed.Joint photographic expert-group (JPEG) is a famous file for images. The Steganalytic of JSteg,JPhide,Outguess 0.13 Outguess 0.2,F5 all use Jpeg Image as the carrier. In this thesis we focused on research of JPEG steganography, Combining with the theory of error-correcting codes and optimization algorithm, we proposed new information hiding methods in DCT domain on the base of the existing methods. Some works have been completed as follows.(1) A steganography method based on Golay code in DCT dominWe study the relations between steganographic algorithms and error-correcting codes,and propose A steganography method based on Golay code in DCT domin。The experimental results show that the proposed method improves the efficiency, achieves a better image quality than F5 algorithm at the same capacity,and keep the property of DCT coefficients histogram,which achieves a better security.(2) JPEG steganography based on Bee Evolutionary Genetic AlgorithmThe proposed method first derives an optimal substitution matrix by BEGA for transforming the secret messages. By means of the substitution strategy, the quality of stego-images is improved. Next, it modifies the standard JPEG quantization table for the purpose of accommodating more secret messages. The transformed messages are then hidden in the cover-image with its DC-to-middle frequency components of the quantized DCT coefficients modified. Finally, a JPEG file is generated through JPEG entropy coding. The experimental results show that the proposed method achieves a better image quality and security than steganographic method without optimal substitution, and it costs less computation time than the genetic algorithm-based information hiding method.(3) Quantization table with k-Fibonacci matrix and human visual system model for JPEG steganographyWe proposed the application of k-Fibonacci matrix and human visual system model on implementing a JPEG steganographic system. The proposed method construct an 8 X 8 k-Fibonacci matrix first and modified it according to human visual system model, the final result is chosen as the new quantization table. Then, the secret message is hidden in the cover-image with its middle-frequency of the quantized DCT coefficients modified. The experimental results show that the proposed method is easy to perform, which not only can hide a larger amount of information in the cover-image compared to other methods, but also can resist category attack effectively.