Studies On The Robustness Of Digital Image Watermarking

Digital watermarking has been proposed as a feasibility strategy to protect copyright. It provides a communication channel to send the watermark signals though complexing the original digital multimedia content and watermark signals without damaging the original digital multimedia content. The watermark signals embedded into the protected digital media as a unique "fingerprint" code. When the copyright issue occurs, the ’fingerprint’ code can be restored out to certificate the ownership. Today, the robust watermark technology has been got the greatly attention as a mainstream technology to protect the copyright. It is mainly divided into two classes:spread spectrum (SS) based watermarking techniques and quantization index modulation (QIM) based watermarking techniques. The angle quantization index modulation (AQIM) and the great gradient watermarking (GDWM) are two dominant QIM methods.This thesis mainly studies robust watermarking in image copyright protection. We propose a difference angle quantization index modulation scheme (DAQIM) based on the angle quantization index modulation (AQIM) and the great gradient watermarking to resist the gain attack. Unlike the original AQIM method, the DAQIM method quantizes the difference of the two angles instead of the angles themselves.1) The main contents of the thesis are described as follows. The main advantage of the DAQIM method is to disperse the interference to the watermark signal from one angle to more angles. Thus, the watermark has a higher ability to resist attacks. We perform a theoretical analysis of the document-to-watermark ratio (DWR) for our proposed method. The results demonstrate that our proposed method outperforms common existing methods in terms of the robustness against various attacks such as the JPEG quantization noise, additive white Gaussian noise (AWGN), cropping effect and mean filtering.2) We use the great gradient vector to embed the watermark. Thus, the watermark has a higher ability to resist gain attacks. We use the standard grayscale image for test. The experimental results show that our proposed method can obtain a lower embedding distortion compared to the AQIM and the gradient direction watermarking (GDWM) methods under the same robustness and payload conditions.