Steganography And Data Hiding Algorithm

The fast developing network doesn't only make the transition and communication of information easier than before, but also puts forward higher requirement to information security which is important to nation security, economic development and individual privacy. Information hiding, as the important embranchment of information security, is becoming the most popular research interest.Information hiding involves mathematics, information theory, computer vision and other computer applied techniques. As the improvement of digital technique, information hiding technique is widely applied into computer communication field. In this thesis, we will mainly research on hiding algorithms of information hiding, which concluded in the following four facets:1. Information hiding technique based on bit-stream recomposition recompose the secret stream and media stream. The recomposed stream still could be used. We discuss different media format and present three information hiding methods of bit-stream recomposition: bit-stream replacing, using the private data of bit-stream and utilizing the error-tolerance function of media decoder. We also present the application of the three information hiding methods into MPEG-1. The information hiding method in PDF file is also discussed in this thesis. The significance of the information hiding technique based on bit-stream recomposition is presenting new idea of inspection of secret information on Internet.2. Digital image information hiding is a subject of the most popular research topic called information hiding. For a given data which we want to keep secret, we just hide it into public digital image and take full advantage of the redundancy of pubic image. In order to protect the secret data, it is required that the difference between the public image and stego-image could not be imperceptible. We present the data hiding algorithm based on harmonic amalgamation. The algorithm based on harmonic amalgamation embeds digital data into a cover image using simple Bezier curve. A message embedded by this approach can be in the form of text, imagery, or any other digital signal. Given appropriate parameters, it can be proved that digital data embedded by this approach could all be recovered correctly by data recovering process with the knowledge of original image. We also present two extraction method of this data hiding algorithm: formula method and look-up method. At last, we present data hiding system based on harmonic amalgamation. In order to increase the security of secret data, the system takes two measures, data encryption and image scrambling, besides embedding secret data into public image.3. Morphological wavelet, as one of non-linear image processing, introduces the erosion and dilation operators of morphology into wavelet which will bring wavelet transform with specialcharacteristics. The extreme-lifting scheme is nonlinear multi-resolution signal decomposition method based on morphological operators. The extreme-lifting scheme preserves local extreme and does not create new extreme in decomposition process. We present an unsymmetrical scheme of the extreme-lifting scheme that can be directly used to decompose two-dimensional image. We also present the proposition of extreme tracing of extreme-lifting scheme and its application in image segmentation.4. As a wavelet transform, morphological wavelet could be applied into data hiding and digital watermarking. For digital watermarking, we use morphological wavelet as a normal wavelet transform and embed the watermark into morphological wavelet transform coefficients. We present another watermarking algorithm based on feature extraction which used morphological wavelet as pre-processing. For data hiding we present morphological wavelet LSB data hiding algorithm.Furthermore, we briefly introduce some kinds of existed steganalysis methods and digital watermark attack methods. At last, we summarize our research work, and discuss those potential research topics in future.