Robust high-definition video watermarking based on self-synchronizing signals against composite distortions

摘要

Abstract. This paper proposes a high-definition video watermarkingnmethod robust against compound geometrical attacks by camcorderncapture as well as common video processing such as frame-rate conver-nsion and transcoding. Unlike traditional watermarking systems using thenoriginal reference pattern, the proposed method exploits the referencenpattern estimated from the watermarked video. Since the reference pat-ntern and watermark pattern are embedded in the same spatial position ofneach frame, the two patterns estimated from the embedded video arenalways in spatial sync. Thus, the watermark information can be simplyndetected without an additional synchronizing step, even if geometricalndistortions happen to the marked video. Also, the problem of misestima-ntion of the reference pattern caused by temporal video clipping is solvednusing the proposed two-pass detector. Extensive experiments prove thatnthe proposed method is robust against various temporal and spatialndistortions. © 2010 Society of Photo-Optical Instrumentation Engineers.nu0001DOI: 10.1117/1.3488053