Active and passive forensic methods for next-generation video content protection

Abstract

Due to remarkable advances in network and display technologies, traditional digital multimedia content that was confined to audio, two-dimensional (2D) images and videos has evolved in various ways. Traditional one-directional content is changed to bi-directional content which encourages the participation of viewers because of the development of smart television and interactive content. Also, one-source, multi-use (OSMU) content, which can be adapted in various platforms and resolutions without any post-processing, comes into the spotlight due to the expansion of cloud computing and N-screen environments. Further, Stereoscopic content or three-dimensional (3D) content has been around since the 1950s, but the market rapidly increased and content poured out at a tremendous rate after the hugely successful movie “Avatar” in the second half of 2009. The evolution of multimedia content is rapidly progressing to the next-generation, as mentioned above. However, the copyright protection techniques for next-generation content are not yet adequate. There is no reported copyright infringement case of next-generation content yet. However, the amount of damage is increased exponentially like the “broken window” effect if the copyright infringement of these next-generation contents occurs. This thesis introduces the active and passive forensic methods to protect the next-generation content. The proposed active forensic method is a stereoscopic video watermarking scheme for protection of DIBR (depth-image-based rendering)-formatted content. The proposed passive forensic method is a screenshot identification scheme due to almost all OSMU content or stereoscopic content is illegally leaked through Internet with their screenshots. The proposed stereoscopic video watermarking scheme suitable for DIBR-formatted content is the technique that the inserted copyright information can be detectable after 2D-3D conversion by exploiting the trait of DIBR. DIBR system renders left and ri...