| The goal of this research has been to design a watermarking scheme, which is robust against common signal processing operations and attacks including data compression. The technique proposed and evaluated in this dissertation leads to a method, which employs an adaptive watermarking system utilizing the wavelet domain basis. The primary emphasis of this dissertation is: (1) The ability to adapt the weighting factors for the embedding process to the statistics of both the watermark and the original image frame, and (2) To generate a self-extracting capability which does not rely on the original watermark or the original data for time varying video. The simulations carried out on several sequences of video imagery and test images indicate that an acceptable quality of recovery at rates of one hundred to one compression is possible. |
