Based On The Analysis Of Image Steganographic With Information Technology Research

Information hiding and steganalysis technology is a hot topic in the field ofinformation security research, and great progress has been achieved in recent years. Byutilizing redundancy of human sensory organs, information hiding realizes hiding onemessage in another without changing basic characteristics and usages of the carriermessage, since the carrier message only shows its external characteristics after finishinginformation hiding. Existing information hiding techniques can be classified as twocategories, the airspace and frequency domain. The former mainly focuses on LSB andpalette based methods, and the latter primarily consists of DCT, DFT and DWT.Steganalysis is used to judge whether certain kind of media file contains hiddeninformation, and estimate the size of the embedded information, or even extract thesecret information embedded inside. Existing steganalysis techniques can be dividedinto dedicated and universal steganalysis (blind detection) methods.This dissertation first introduces the basic principle of information hiding andthen concentrates on the universal theoretical model of image hiding, the classificationand features of information hiding technology, in addition, the concept, basic principleand classification of steganography are also studied. Afterwards, this dissertation looksdeeply into basic theories of the classical information hiding algorithms by analyzing intwo directions respectively, the airspace and frequency-domain, and goes deeply intotheir basic theories along with the classical algorithms of steganography in professionaland universal way. And then this dissertation concentrates on the storage features ofprevailing image format, and gives an outlook of the application of information hidingtechnology.Moreover, two new information hiding techniques based on the image airspaceare proposed. The first one is called HSB algorithm, which utilizes the fact that imagecharacteristics are established on its maximum bit-planes. HSB algorithm caneffectively increase the capacity of the secret information by hiding the biggestbit-planes of an image in less demanding circumstances. And the second one is EOBoBalgorithm, it is theoretically based on characteristics of the gray histogram and plane. By calculating the secret information and bits in the second high plane with bit XOR,EOBoB effectively reduces changes in the histogram and the statistic features of thelowest effective bits. Relative experiments reveal that the number of adjacent pixelswith differences equaling0and1changes dramatically after hiding information, thusthis dissertation puts forward the S detection algorithm. Finally, a PS detectionalgorithm based on the24-bit BMP image is also proposed by statistically analyzingchanges in the sum of XOR values among eight adjacent pixels after hiding information,and further tests prove the correctness and effectiveness of the new algorithms.