The Outline Of Wave Digital Image Watermarking Algorithm Based On Fractal Theory Research

In recent years, digital watermarking technology caused widespread concerns. It has bothacademic and commercial values. When we enter the digital age, destruction of information isinevitable. In the hightly developed network environment, how to protect the legitimate rights andinterests of everyone becomes a pressing issue.This thesis describes two transforming domain watermark algorithms basing on fractal. Thefirst one is the wavelet domain algorithm, In the watermark embedding process, we have takenadvantage of the the hearst matrix generated by the fractional brownian motion model, itsuccessfully resolved the problem of watermarking visibility due to human eye’ sensibility to lowwavelet frequency. The other one is the contourlet algorithm.To begin with, it analyzes the fractalcharacteristics of the host picture and extracts the fractal dimension. Fractal dimension can reflectthe complexity of the host the image such as contour and texture, a functional relationship isestablished with the embedding strength. So the adaptive watermark strength can be achieved. Inorder to enhance the security of the watermark, both algorithms are combined with two differentchaotic encryption methods.In this thesis, we completed a large number of comparative experiments and calculated thevarious values. The results showed the algorithms are very good to solve the contradiction betweenthe transparency and robustness of the watermark. Also both of them have strong ability to resistconventional signal processing means such as geometric attacks, filtering attacks, noise attacks, andcompression attacks. The proposed schemes demonstrate very good performance.