An Image Digital Watermarking Algorithm Based On Discrete Wavelet Transform

Information hiding is a kind of method to protect products copyright. With the rapid development of the digital and network techniques, the information hiding technology has become a hotspot in the current information science field. Digital watermarking is a main part of information hiding technology.Through embedding the invisible information into digital media, and then tracking the use of image by checking the hiding information, this kind of technology can implement information hiding, storage, copyright protection and so on.Unobtrusiveness and robustness are the two elementary characteristics of digital watermarking technology. Because mankind eyes are more sensitive to low frequency parts than to high frequency parts, the methods of embedding digital watermarking into the low frequency parts may make the quality of digital image worse and decrease the unobtrusiveness of the watermarking. The better unobtrusiveness of watermarking can be obtained by using the methods of embedding digital watermarking into the high frequency parts.However, the watermarking embedded is easy to be destroyed by some signal processes such as image lossy compression. The goal of this thesis is to design a better watermarking embedding method which has a good characteristic of unobtrusiveness and robustness.Based on wavelet domain theory, the thesis analyzed and researched digital watermarking technology from following aspects:(1)Researched the concept, principle, watermarking system model, common watermarking algorithm and attack methods of digital watermarking. Systematically explicated and realized the image scrambling technology,which is one way of watermarking pretreatment.(2)Researched the wavelet multiresolution analysis thinking, mallat algorithm, and gave the filter's realization solution of mallat algorithm.Analysed and realized the digital watermarking's digital image decomposition technique based on wavelet transformation. Vertified defects in terms of unobtrusiveness and robustness with the low-frequency and high-frequency scheme in the digital watermarking,and proposed some improvement thoughts.(3)The thesis improved these two watermarking embedding programme and proposes a watermarking embedding method in the low-frequency part of wavelet domain. Seven sub-bands of mid and low frequency were obtained after the wavelet transform on the image, on account of the local maximum principle, the algorithm chose the potential embedding coefficients, within the range of the seven sub-bands. In the end, applying the multiplicative update, the watermarking was embedded into the selected potential embedded coefficients, and the embed intension self-adaptively was adjusted, based on the characteristic of frequency mask of the HVS.The thesis analysed the performance of the algorithm through a large number of attack experiments.The results of experiment show that the method which paper provided could bring the compromise between the watermarking robustness and unobtrusiveness. Embedded watermarking image quality had not decreased significantly and the watermarking embedded was effective to be extracted after the processes such as a lossy compression, median filter, tailoring and additional noise and so on.