Research On Digital Image Watermark Technology Based On Chaotic Encryption And Wavelet Transform

Internet has become an important channel for the exchange of information in modernsociety, and the following information security problems have increasingly attracted people’sattention. As an information carrier, image is much more acceptable and used more widelythan text because of its large amount of data and high redundancy, so the protection of imageinformation has increasingly become a hot topic at home and abroad. Based on this, imageencryption technology and image information hiding technique are studied in this thesis, animage encryption algorithm is proposed, and on the basis of which an watermarking algorithmis designed, finally the two methods are simulated and analyzed.This thesis mainly makes discussion and research on the following aspects:(1) The chaotic characteristics of Logistic map and Skew Tent map are analyzed. Thebifurcation diagram and traversal graph of the two chaotic maps are drawn, simulation resultsshow that Skew Tent map has better ergodic performance and greater range of parametersthan Logistic map, so choosing Skew Tent map to scramble image has better scramblingeffect.(2) Aiming at the problems that Logistic map has more concentrated iteration points andpoorer traversal performance when used for image encryption, an improved image encryptionalgorithm based on Skew Tent map and DNA (Deoxyribonucleic Acid) theory is proposed.Firstly this algorithm uses Skew Tent map to produce two chaotic sequences to scramble thelocations of pixels, then the initial scrambled image is coded by DNA rules to make it a DNAsequence, then Skew Tent map is used again to produce one chaotic sequence to scramble theDNA sequence, finally the encryption image is obtained by DNA inverse transform. Thealgorithm is simulated and analyzed from two aspects of security and scrambling degree andcompared with some existing them. Simulation results show that the algorithm has goodencryption features.(3) Aiming at the fact that watermarking algorithm requires both robustness and in-visibility, an image watermarking algorithm based on chaotic encryption and wavelettransformation is proposed. Firstly this algorithm uses Skew Tent map and DNA encoding toscramble watermark, then the carrier image is decomposed by block wavelet transform, thenSVD (Singular Value Decomposition) is made to the low-frequency part of image block, andchaotic genetic algorithm is used to optimize embedding factor, the scrambled watermark isembedded to the singular value of low-frequency part, the watermarking image is obtained byinverse wavelet transform. Image filtering, image compression and other treatment are madeto the watermarking image, and algorithm characteristics are analyzed from two aspects of peak signal to noise ratio and normalized correlation coefficient. Simulation results show thatthe algorithm has better robustness and invisibility.Some conclusions can be got through theoretical research and simulation analysis of thethesis, using chaos theory and DNA encoding to image encryption not only achieves goodresults, but also can it have a better pre-scrambling effect for watermark embedding; applyingchaos theory and wavelet transform to watermark embedding process and taking into accountthe impact of embedding location and embedding factor in the process, which makeswatermark have better robustness and invisibility after embedding.