SmartMark Technology

Advanced Video Piracy Protection Technology

Overview
SmartMark is the "digital watermark" technology that makes MediaSentinel a powerful tool in the fight against piracy. SmartMarks are digitally encoded forensic data invisibly embedded into a video stream. To create a SmartMark, MediaSentinel performs a real-time digital video content preprocessing that determines a level of a content (image/video) dynamics (static scenery, medium-to-fast motion) and the distribution of frequencies in both luminent and chrominent digital video components as main criteria for a dynamic adjustment of key SmartMark parameters. The fundamental objective is to preserve total visual imperceptibility (transparency) while increasing SmartMark robustness and sustainability to potential piracy attacks.

Watch Videotest
SmartMarks In Action

Description
The level of a video content dynamics is also called the “density of information”. The content of every video and image can be characterized by its distribution of the density of information. A video preprocessing on a frame-by-frame basis determines the density distribution. Large image areas with plain colors or limited variations of colors per area (blue sky, water, plain color wall etc.) become “discovered” first during the initial step of fast video (frame) content pre-scanning session. The next step automatically adjusts and distributes key SmartMark parameters (amplitude, bits per frame) to fit the “low dynamics” frame area(s). The density of inserted information—the SmartMark itself—is minimal, resulting in no visible degradation. On the other hand, if the whole frame or a certain portion(s) of a frame is rated high dynamics area(s), with both luminent and chrominent components of a digital video carrying moderate to large variation of colors and intensity components, key SmartMark insertion algorithm parameters values are changed (increased) in order to maximize the probability of the correct detection by the SmartMark Reader.

Algorithm
The SmartMark Algorithm combines relevant methods for 2-dimensional signal processing (digital image processing) such as:

- Dispersion analysis,
- Two-dimensional filtering
- Factor analysis.

SmartMark searches for and detects “high risk area(s)” within the video frame and characterizes it by objective processing parameters such as

- Statistical spectral distribution,
- Size,
- Geometry

Using this information the algorithm calculates and dynamically optimizes key SmartMark insertion parameters values for the SmartMark insertion process. This optimization occurs in real-time.

Because the data pattern changes from frame to frame and the generating mechanism is itself randomized from frame to frame, the possibility of successfully removing the data from every frame in an entire video is remote. The SmartMark algorithm is numerically efficient enough to be implemented in real-time when the video is being rendered.

SmartMarks of digital video data are created by inserting the encoded data into individual color channels or the intensity channel of a digitized content. The technology is based on time-to-frequency domain transform (Fourier spectrum) with spectral lines inserted in the perceptually insignificant portions of the spectrum of the individual video frames. Inphase and quadrature components or sinusoids may be encoded to two chroma channels to provide shift-invariant detection of the spectral lines. A cryptographically secure random number generator is used to select signaling frequencies and phases of the inserted spectral lines to defeat overwriting attacks. Specifically, a SmartMark is inserted into the perceptually most insignificant components of the data to be virtually imperceptible. A narrow band signal representing the SmartMark is placed in a wideband channel that is the data. The SmartMark can be embedded in real time simultaneously with the authorized video distribution event.

To make SmartMarks resistant to overwriting, a spread-spectrum concept is employed. The frequencies are selected at random from a larger set than necessary. This leaves a lot of “silent” bandwidth in the SmartMark spectrum. If an attacker wishes to cover up the SmartMark, he must cover up the entire available spectrum. Such an attack will seriously degrade the image quality.

The proliferation of digital content demands a security system that works across all delivery platforms and viewing formats. Although SmartMarks do not prevent unauthorized duplications, they deter piracy by preserving in every copy, specific, unique encrypted identifiers (SmartMarks) associated with the original video delivery, allowing pirated copies to be traced back to the source.

SmartMarks intelligently respond to image characteristics to ensure invisibility even in extreme situations.

SmartMarks do not interfere with the quality of the video content or its enjoyment by consumers nor do they raise any compatibility issues within mainstream video player environments.

Survivability
SmartMarks survive video processing attacks, such as digital-to-analog-digital conversions (including multiple D-A/A-D conversions), re-sampling and requantization (including dithering and recompression) and common signal enhancements to image contrast and color.

SmartMarks survive “StirMark” attacks, including resizing, letterboxing, aperture control, low-pass filtering and anti-aliasing, frame-swapping, compression, scaling, cropping, overwriting, the addition of noise and other transformations.

SmartMarks survive attempts at sync separation, the cutting off or damaging of the horizontal and vertical sync signals that are embedded into a composite (analog) video as well as into a digital video stream.

SmartMarks survive format conversion, the changing of frequencies and spatial resolution among NTSC or PAL, into another and vice versa.

SmartMarks are embedded in digital video frames by adding complex exponential or sinusoidal signals to the chroma or intensity information in each frame. Chroma data consists of two channels for each pixel; intensity consists of one channel for each pixel. This identifying information can be used to trace the origin of pirated video data. Each pixel in the frame is represented by a triple consisting of a red, green, and blue component. This triple is linearly related to intensity, Y, and 2 chroma components, which is called the “YUV” decomposition. The channel designated “Y” is the intensity, and the “U” and “V” components contain the color information. SmartMark core technology uses two arbitrary chroma components. They can be called “U” and “V”. The SmartMark method adds small increments to “U” and “V”. These increments are recovered when the SmartMark is read and can be interpreted as the real and imaginary parts of a two-dimensional complex exponential signal. The components “U” and “V” can be constructed to enable SmartMark hiding, transfer of the SmartMark through any number of transformations and compressions, and computational efficiency. With complex-valued color SmartMarks, positive and negative frequencies in the horizontal and vertical dimensions are used.

Because “U” and “V” are orthogonal, the increments can be recovered as the SmartMark is “read”. There is no “crosstalk” between the two increments. Thus, each pixel can be used to deliver two small increments without changing the intensity of the pixel.

Last Updated ( Saturday, 18 November 2006 )