A JND-based saliency map fusion method for digital video watermarking

Growing popularity of video based applications such as Internet multimedia have a demand for copyright protection and authentication. Digital video watermarking can be applied for this task. Traditional video watermarking methods have been limited to signal processing field and pay less attention on Human Visual System(HVS) model. In this paper, we propose a novel digital video watermarking framework for finding more efficient embedding positions. The key idea is to leverage Just Noticeable Distortion(JND) to serve as visual sensitivity descriptor, thus most attractive areas could be extracted from video frames. Specifically, We use JND as thresholds to describe motion saliency area and fuse with static saliency area to better represent keyframes. Experimental results demonstrate the effectiveness of our proposed method compared with traditional block-DCT watermarking method. Meanwhile we give statistical analysis of keyframe intervals in experimental section.