A Self-Adaptive Video Dual Watermarking Based on the Motion Characteristic and Geometric Invariant for Ubiquitous Multimedia

Abstract

In ubiquitous multimedia applications, how to protect the digital copyright for multimedia data has been a difficult task. In this paper, a novel self-adaptive video dual watermarking, which is combined the motion characteristics detection with the geometric invariant of Scale-invariant feature transform (SIFT) is proposed. For each frame, the motion characteristics are calculated as the maximum of embedding strength. Moreover, the maximum of embedding strength are firstly embedded into the middle and high frequency bands in discrete wavelet transform (DWT) domain as the first watermarking signals. Then the watermarked frame is used to obtain the feature points with the invariant performance of SIFT descriptors. These feature points, as the other watermark, are registered and stored in IPR database as zero-watermarking. In the process of watermarking extraction, to each frame, the SIFT feature points are acquired through zero-watermarking. By using the scale feature and coordinate relationship of the SIFT feature points, the video frames are detected and corrected to makes the extraction of the watermark information is synchronized. Experiments indicate that using Motion Characteristic and geometric invariant of SIFT can significantly improve the watermarking imperceptibility, effectively resist common geometrical attacks for video watermarking and consequently achieve higher robustness.