Robust Image Watermarking in the Print-Cam Process.

Nowadays with the popularity of smartphones, it is increasingly valuable to capture side information in print media through phone cameras. The print-cam process produces serious attacks during capturing images, especially geometric distortions. In this paper, we propose a robust image watermarking method for the print-cam process, which is also evaluated on some different smartphones. In the embedding process, the method uses watermarking based on Discrete Cosine Transform (DCT). We introduce Scale-Invariant Feature Transform (SIFT), the local feature point detection techniques, to restore the geometric distortions, and then the reverse process of embedding is carried out to extract the watermark. The results show that the proposed method is considerably robust in the print-cam process. Meanwhile, the method is suitable for a wide range of smartphones and acceptable for real-life applications such as ID card verification, anti-counterfeit traceability, and medicine authentication.