PTAB

IPR2017-01675

LG Electronics Inc v. Broadcom Corp

Log In
Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Graphics Display System with Anti-flutter Filtering and Vertical Scaling Feature
  • Brief Description: The ’104 patent describes a graphics display system and methods for blending graphics and video images for simultaneous display. The core of the invention is a claimed two-step process: (1) blending a plurality of graphics images together using their associated alpha (transparency) values to generate an intermediate “blended graphics image,” and (2) blending that intermediate image with a video image using the original alpha values or a composite alpha value derived from them.

3. Grounds for Unpatentability

Ground 1: Obviousness over Gloudemans and Blinn - Claims 1, 10-11, 16-17, and 22 are obvious over Gloudemans in view of Blinn.

  • Prior Art Relied Upon: Gloudemans (Patent 6,266,100) and Blinn (an IEEE article, "Compositing Part 1: Theory," Sep. 1994).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Gloudemans disclosed a system for enhancing live video presentations, such as football games, by overlaying graphics (e.g., a first-down line, logos) onto a video feed. This system included the necessary circuits, processors, and memory recited in the independent claims. Gloudemans taught creating graphics and associated alpha values and sending them to a "keyer" to blend with the video.
    • Motivation to Combine: Petitioner contended that while Gloudemans showed blending single graphics, it provided examples where multiple graphics (e.g., a first-down line and a score box) would inevitably overlap. To improve the viewer experience in these predictable situations, a Person of Ordinary Skill in the Art (POSITA) would be motivated to blend the overlapping graphics together first before overlaying them onto the video. Blinn taught that alpha blending is associative, meaning multiple foreground images could be pre-blended into an intermediate image with a new, correctly derived composite alpha value. A POSITA would combine Blinn’s well-known compositing theory with Gloudemans’ system to efficiently handle overlapping graphics, reduce processing latency in a live broadcast, and simplify processing, as only a single blended graphic would need to be scaled and synchronized with the video.
    • Expectation of Success: A POSITA would have a high expectation of success because Blinn’s teachings on the associative property of alpha blending provided a clear and mathematically predictable method for pre-compositing graphics layers.

Ground 2: Obviousness over Myhrvold and Blinn - Claims 1, 10-11, 16-17, and 22 are obvious over Myhrvold in view of Blinn.

  • Prior Art Relied Upon: Myhrvold (Patent 5,867,166) and Blinn (an IEEE article, "Compositing Part 1: Theory," Sep. 1994).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner asserted that Myhrvold taught an image processing system for real-time rendering that composites multiple graphics layers (called “gsprites”) to create a final image. Myhrvold explicitly disclosed blending multiple gsprites together pixel-by-pixel based on their individual alpha values and storing the blended result. Crucially, Myhrvold’s system accumulated opacity for each pixel in an “alpha buffer,” which Petitioner argued was functionally the same as the “composite alpha value” of the ’104 patent. Therefore, Myhrvold taught the first step of the challenged claims: blending multiple graphics to create a blended image and a composite alpha value.
    • Motivation to Combine: Myhrvold disclosed that its system could combine video and graphics. Petitioner argued it would have been obvious to apply Blinn’s teaching to Myhrvold’s system to perform the second step of the claimed process. Blinn taught that an alpha-blended foreground image (like Myhrvold’s composited gsprites) could be subsequently overlaid onto a background image using a derived alpha value. A POSITA would be motivated to use the composite alpha value already stored in Myhrvold’s alpha buffer to blend the composited gsprites with a video background. This would be a straightforward application of Blinn's technique to Myhrvold's existing architecture to achieve the predictable result of combining a complex graphic with video while maintaining correct transparency.
    • Expectation of Success: The combination was argued to be predictable because Myhrvold already generated the necessary components (a blended graphic and a composite alpha value), and Blinn provided the established method for using those components in a subsequent blend with a background.

4. Key Technical Contentions (Beyond Claim Construction)

  • Associative Property of Alpha Blending: The central technical contention underpinning both grounds was that the associative property of alpha blending was a fundamental and well-understood principle in computer graphics prior to the patent’s filing date. Petitioner argued this principle, explicitly detailed in Blinn, dictates that the order of blending operations can be changed without affecting the final result (i.e., (A over B) over C = A over (B over C)). Consequently, the ’104 patent’s “two-step” process was not an inventive concept but merely an obvious design choice among other known options (e.g., blending each graphic individually with the video).

5. Relief Requested

  • Petitioner requested institution of an inter partes review and cancellation of claims 1, 10, 11, 16, 17, and 22 of the ’104 patent as unpatentable under 35 U.S.C. §103.