Robust and blind video watermarking against online sharing platforms

Area of Science:

• Digital watermarking
• Video copyright protection
• Multimedia security

Background:

• The proliferation of online video sharing necessitates robust copyright protection methods.
• Existing block-based watermarking techniques are vulnerable to geometric distortions, hindering accurate watermark extraction.
• The increasing prevalence of video sharing platforms poses challenges for traditional watermarking schemes.

Purpose of the Study:

• To propose a robust and blind video watermarking scheme resilient to online sharing platform manipulations.
• To enhance the accuracy of watermark localization and extraction.
• To improve the balance between watermark invisibility and robustness.

Main Methods:

• Preprocessing video using Dual-Tree Complex Wavelet Transform (DTCWT) to mitigate non-geometric signal interference.
• Dividing the low-frequency domain into concentric rings to establish spatial stability regions for synchronized extraction.
• Optimizing watermark embedding by analyzing and adjusting the energy relationship between adjacent ring subbands using Discrete Cosine Transform (DCT).

Main Results:

• The proposed watermarking scheme demonstrates significant robustness against common video processing operations.
• The method exhibits strong resilience to distortions introduced by video online sharing platforms.
• Experimental results validate the effectiveness of the DTCWT and ring-based approach for watermark security.

Conclusions:

• The developed video watermarking scheme offers a reliable solution for copyright protection in the context of online video sharing.
• The integration of DTCWT and a novel ring-based embedding strategy enhances robustness and accuracy.
• The scheme effectively addresses the limitations of existing methods, providing improved security for digital video content.