A new efficient anaglyph 3D image and video watermarking technique minimizing generation deficiencies

3D Anaglyph system is among the most popular 3D displaying techniques thanks to its simplicity and the cheap glasses that it uses. Anaglyph generation and watermarking are two essential techniques that attracted researchers in 3D anaglyph domain where several techniques have been proposed. However, most of the previous anaglyph watermarking studies focused on the robustness and the visual difference between original and marked content and they have not considered the three deficiencies caused by the generation step, which are the distortion of colors, the retinal rivalry and the ghosting effect. In this paper, we propose the first watermarking technique that protect 3D anaglyph content before its transmission by embedding the signature simultaneously with generation step. In this technique, three signatures were embedded before, during and after the generation process using different domains to obtain robustness to several manipulations, especially against malicious attacks. Moreover, the chosen generation process avoids generation deficiencies and allows obtaining high visual quality of the marked content. The experimental results illustrate robustness against attacks such as compression and collusion where the minimum value of NC is close to 0.7 and the maximum value of BER is close to 0.2. Besides, the suggested technique provides high invisibility where PSNR and SSIM values are respectively close to 58 and 0.9, and it minimizes the generation deficiencies.