Improving robustness of visible image watermarks

Abstract

The visible watermarking is an important intellectual property rights (IPR) protection technique for digital images. In some circumstances such as multimedia contents used in learning websites or digital libraries, digital images have to be released but illegal reproductions of them are prohibited. Digital images embedded with visible watermarks contain perceptible but unobtrusive patterns. The embedded patterns should be difficult to remove unless expensive human labours are involved. Recently, Huang and Wu have proposed an efficient attacking scheme against visible watermarks. Based on the attacking scheme, the watermark patterns can be almost completely eliminated and a perceptually satisfying recovered image can be obtained by means of limited human interventions. In the present article, a novel scheme is proposed to improve the robustness of visible watermarks. For counterattacking the removal of the visible watermark pattern, the significant information of watermark patterns is extracted and embedded in the protected image imperceptibly by the information hiding techniques. In addition, to specify the positions of destroyed areas of watermarks, a fragile watermarking technique is designed and applied in the proposed scheme. Simulation results demonstrate that the proposed scheme is robust to the up to date attacks.

Keywords:

• FRAGILE WATERMARKING
• INFORMATION HIDING
• VISIBLE WATERMARKING
• WATERMARK ATTACKS