Robust Watermarking for Scalable Coded Video Streaming
Funding: EPSRC Dorothy Hodgkin Award
Digital watermarking involves imperceptibly altering a digital media work to embed a message about the work. Watermarking is used in a variety of applications including owner identification, content authentication, broadcast monitoring, transaction tracking and in media digital rights management (DRM) applications within content supply chain. Robustness, which is the ability to detect the watermark after common signal processing operations, is one of the useful watermarking properties required for many applications.
Scalable coding is an emerging efficient way of representing media content to facilitate seamless delivery of multimedia content along the supply chain from content providers to the consumer. In scalable coding, high resolution content is encoded (shown as "E" in the figure) to the highest visual quality and the bit streams are stored in central servers. In order to cater various display terminals and to stream along various channels, new bit streams with lower bit rates and/or lower resolutions are extracted (shown as "X" in the figure) according to the channel, device and usage requirements. These bit streams can be re-adapted at various nodes of streaming and finally decoded (shown as "D" in the figure) for playback at terminal displays.
In this project, such content adaptations of scalable coded media along supply chains are modelled and these models will be used to develop new watermark embedding and detection methods to improve the robustness of watermarks to such content adaptation attacks.