Screen-Camera robust watermarking using Arnold Transform and Double-Density Dual-Tree Discrete Wavelet Transform

Digital photo acquisition, manipulation, and redistribution are ubiquitous because of digital globalization. Digital watermarking provides a comparative solution for leak information tracking and retrieval problems and improves authentication, security, and copyright protection in today’s digital globalization. This paper presents a watermarking scheme joined with the Arnold encryption technique for Screen-Camera (SC) rebroadcast attacks. This research aims to extract watermark information from the stego (or watermarked) content even after a rebroadcast attack from the LCD screens. The technique is resilient to lens distortion, light source imbalance, and aliasing artifacts. To resist the geometric distortions caused by the recapturing process, we propose a cascaded Double-Density Dual-Tree Discrete Wavelet Transform (DDD-DWT) domain and Singular Value Decomposition (SVD) based algorithm. However, most SVD-based watermarking techniques are prone to false and missed alarm rates and cause security issues. On that account, this paper presents the Arnold Transform-based watermark encryption technique using a unique security key for authentication on the decoder side. The evaluated results showed that the scheme achieved a high Peak Signal-to-Noise Ratio (PSNR) value of 64 dB, an Structural Similarity Index (SSIM) value of about 0.99, and robustness to several geometrical, signal processing, compression, and recapturing attacks.