Enhanced Video Watermarking Using Hybrid 2D-DWT, SVD, and Paillier Homomorphic Encryption
Khushboo Chhikara, Manoj Kumar · 2025
Video watermarking is essential for protecting intellectual property rights and maintaining content authenticity. Existing video watermarking techniques often face challenges related to robustness, imperceptibility, and security. To overcome these challenges, this paper presents a robust and secure watermarking framework using advanced embedding techniques and encryption methods for enhanced security. Initially, the video file is divided into frames for processing. After that, 2D-Discrete Wavelet Transform (2DDWT) is employed for converting processed frames into subbands. The LL sub-band is selected for embedding because it's high robustness and suitable for imperceptible watermarking. Following that, the chosen sub-band is subjected to Singular Value Decomposition (SVD). The watermark image is encrypted using Paillier Homomorphic Encryption (PHE) to ensure confidentiality and security and then concealed in the modified singular values (S). Later, the watermark is extracted by reversing the embedding process and retrieved from the singular values of the sub-band. The proposed method is implemented using Python. The suggested watermarking framework is assessed utilizing various metrics such as PSNR, SSIM, and embedding capacity, and its performance is validated against four different attack scenarios. According to experimental findings, proposed framework outperforms current methods in watermarking applications by ensuring high security, robustness, and quality preservation.