Analysis of watermarking systems in the frequency domain and its application to design of robust watermarking systems

This paper aims to design robust watermarking systems in the frequency domain. We first present the general model of watermark embedding and extracting processes and carry out their analysis. Then we examine the robustness of the watermarking system against common image processing and clarify the reason why detection errors occur in the watermark extracting process. Based on the result, we improve the watermark extracting process and design the robust watermarking systems. The improvement is accomplished using deconvolution filter and neural network techniques. Numerical experiments using the DCT-based watermarking system show good performance as we expected.