Forensic watermarking or digital watermarking is the process of embedding an imperceptible sequence of code or a set of characters into a document, video, audio, or program which enables the unique identification of the content creator and its authorized user. A forensic watermark is embedded into the video content during the encoding process and cannot be easily manipulated later either by reencoding, resizing, or scaling. This ensures content protection while not interfering with the viewing experience. With the help of forensic watermarking, content ownership can be verified and the source of piracy be easily identified by extracting the watermark payload using a specialist cloud-based service, which can work effectively even on low-grade and recompressed videos.
Forensic video watermarking work in the following ways:
- Streaming session or user account identification: This is used for unicast delivery and video-on-demand where the ID of the user account or streaming session is embedded into the video during playback. In the case of live streaming, illegal streaming sites are monitored and when a user ID is retrieved, the streaming session can be taken down or the user account shut down.
- Device Identification: In this case, the device can insert its unique identification before showing the content. For example, the serial number or smartcard ID of a set-top box can help trace the user owning the subscription or device.
- Recipient Identification: This is used when the content owner shares the content with third parties, such as sub-contractors, redistributors, or agencies for dubbing or subtitling. The sender can mark each copy with a unique ID associated with the recipient along with a legal warning to prevent illegal sharing.
- Digital cinema: When movies are played in theaters, the projected video is embedded with a projector ID and time code, which when accompanied by other enforcement processes, is effective in managing screen-capture piracy.
DRM video protection, combined with forensic watermarking, offers the next level of protection from video piracy. Watermarks can be embedded at the server-side, on an edge server, at the client-side, or a combination of these. At the client-side, the logic is implemented at the firmware or SDK level and the information is generated as a randomized ID at the server-side. When a hybrid approach is taken, the server still pre-processes the content to build different versions and the watermark is managed at the edge servers or on the client-side. With OTT platforms focusing on the user experience and cost reductions, client-side and hybrid approaches are more frequently adopted.