Digital Video Watermarking in Lower Variance Sub-bands of Curvelet Transform

Abstract

The protection of multimedia data is becoming more important in video watermarking. The advancement of digital watermarking can help to overcome this difficulty. The major challenges in watermark techniques are robustness and invisibility.  This research proposes an invisible and resilient video watermarking system. It uses a transform domain in embedding the watermark. Only few frames are selected for watermarking using intensity-based seed generation. The structural and textural properties of the selected frames are studied using phase congruency and entropy. Then curvlet transform is applied from which lower variance sub-bands are evaluated for future stages during the embedding phase. The proposed method is tested on three challenging videos and the performance is evaluated using Peak Signal to Noise Ratio (PSNR), Mean Square Error (MSE) and Normalized Correlation Coefficient(NCC).The efficiency of the proposed method is proved by comparing it with other transform domains. The proposed method outperforms the existing methods by a wide margin