Study on robust image watermarking techniques

Abstract

The recent growth in computer networks, and more specifically, the World Wide Web, has allowed multimedia data such as images to be easily distributed over the Internet. However, many publishers may be reluctant to show their work on the Internet due to a lack of security. Images can be easily duplicated and distributed without the owner’s consent. Robust digital image watermark have been proposed as a way to tackle this issue. An effective robust digital watermark should be perceptually invisible to prevent obstruction of the original image. It should be statistically invisible to prevent detection, and it should also be robust to many image manipulations, such as filtering, resize, addition of random noise and JPEG compression. The report is to introduce a number of different techniques being used for robust digital image watermarking such as LSB (Least Significant Bit) Modification, Comparison-Based Correlation using block processing in spatial domain and DCT (Discrete-Cosine-Transform) based method in frequency domain. The algorithm of different image watermarking techniques is included. And matlab is tool for designing and implementing robust digital image watermarking. All source codes, experiment results and robustness evaluation are presented. The aim of the project is to use different techniques for implementing robust digital image watermarking and evaluating the robustness. In addition, the limitation and possible improvement for different techniques are recommended for future improvement.