How video watermark embedding process secures premium OTT content against unauthorized use

Over-the-top (OTT) platforms require multiple strategies to tackle piracy, which leaks their revenue to illegal players. Hollywood studios stress on the need to not just contain piracy through digital rights management (DRM) technology but also want copyright information to be embedded in video files in the eventuality of a successful piracy attack. OTT content has, thus, come to be protected using a combination of DRM technology and embedding forensic watermarks in video frames of premium content.

Both these tasks can be accomplished by the tools provided by multi-DRM vendors. The SDK of such a vendor embeds binary strings in video files which contain copyright information and user information. The embedding algorithm provided by the multi-DRM vendor supplements the DRM video protection with invisible and robust watermarks that stay tamper proof even if the video file undergoes multiple rounds of manipulation, like format change.

Depending upon which process the algorithm employs, video watermarking can be divided into the following three categories:

• Original-video-based watermarking algorithms: As the name suggests, this type of algorithm treats the original video before compression. The original video is interpreted as a series of still images (video frames) which are bound together in a sequence by a frame rate. Watermark is embedded to the original video and then it is compressed.
• Video watermarking algorithms in the encoding process: In this process, the watermark is embedded in the video during the compression coding process. The algorithm modifies redundant spaces in the file to insert watermarks and has little effect on code rate of video streams. This method allows for watermark extraction in real time during the decoding process if required modification is made to the encoder.
• Video watermarking algorithms after compression: A compressed video stream still has redundant spaces which can be used to embed video watermarks after the compression process is over. Unlike in the previous method, this method keeps the algorithm and the codec separate, thus improving efficiency.

If a premium video file leaks, content producers use a watermark extraction algorithm to do the reverse of the watermark embedding process. The nature of the extraction algorithm depends on which embedding method is used to insert watermarks.