Forensic watermarking has become a technology of prime importance with the proliferation of 4K and UHD OTT and VoD content. It arms content creators and distributors with a security suite robust enough to handle multifarious piracy attacks, including simulated attacks on the watermarks.

DRM protected content does come with a certain level of security, but watermarking technology makes it more robust. The imperceptible watermarks hidden at random places in video frames throughout the content can trace back the copyright of pirated content and help source identification. The watermark embeds on top of the DRM protected content in SDK without additional performance lags. The SDK facilitates an interactive interface for the transcoding operation, which decodes each video frame from the media content. Each input frame fed into the SDK outputs watermarked frames with the same set of parameters like video width, height, and pixel ratio. The transcoder then encodes these watermarked frames into the output media for distribution.

The video watermarking technology is effective irrespective of codec format used and can support both online and offline modes of watermarking. An efficient process of watermarking does not induce any visible performance lags and maintains the integrity of the original video stream, with a similar count of output video frames as the original. It is also robust enough to handle various types of simulated manipulation attacks aimed at destroying the watermark.

A watermark’s robustness is determined by its behavior when subjected to a series of simulated attacks and the limit beyond which it becomes unrecoverable.

The simulated attacks involve transformations of various kinds, including geometrical deformations, recompression, merge, and trim attacks. These techniques include upscaling and downscaling video frames, horizontal and vertical flips, different degrees of rotation, tampering with framerates, HDMI capture, screencasting, video capture using devices, insertion of black bars, bit rate reduction, etc.

The recoverability of the watermark against these simulated attacks should be quick enough to make it an ideal candidate for withering attacks intended for illegal capture and distribution of premium content.

Testers, thus, use a comprehensive set of simulated attack types on video assets with varying intensities to examine the robustness and commercial viability of the video watermark against real-time pirate attacks.