The Application Research Of Information Hiding On E-business

E-business as one of the most popular application of the modern computer network, its development is restricted by the information security. In the same time, information hiding technology as one of the focus in information security field can be widely used to E-business with its particular merit. This paper mainly researches the application of information hiding in E-business. The paper introduce various technology of steganography and digital watermarking aim at insuring confidentiality, integrality, anti-fake, non-repudiation and exclusive of information in E-business.The major work involved is as follows:1. The paper proposes a modified algorithm of LSB based on logistic chaotic sequence, which can improve the demerit of the simple algorithm of LSB replacement and the algorithm of LSB matching. The first one usually lead to asymmetry in statistics and abnormal phenomena in histogram of the Stego-images, the latter is non-blind extraction algorithm. The algorithm of the paper use two logistic chaotic sequences, one used to scramble the embedden image, the other used to randomly fix the location which will be embedded in the lowest bit of the cover image. This algorithm can avoid the unbalanced of the 0 and 1 in LSB and it belongs to the algorithm of blind extraction. The experimental results demonstrate that this method is invisible and safe.2. A new algorithm based on discrete wavelet transform (DWT) is presented. The watermark is embedded into a cover-image twice according to this algorithm. At the vary start, the DWT was performed on cover image. Then the watermark embedding location was identified which is the texture and smooth area except the strong edges area in accordance with the value of the high frequency sub-image HH,HL,LH. First, the binary watermark was embedded into the sub-image of HH by exchanging the adjacent wavelet coefficients in the watermark embedding location for the sake of improving the transparency and reducing the probability of interference. Second, the binary watermark was embedded into the sub-image of LL by the same way in order to the watermark not be easily lost and the watermarked image was not degraded. In addition, the final extracted watermark is the result of the two watermarks which were extracted from HH and LL sub-image and then were added and averaged. The experimental results demonstrate that the algorithm is both transparent and robust and can effectively guarantee the anti-fake, exclusive and non-repudiation of information.