PTAB

IPR2017-00336

Samsung Electronics Co Ltd v. Image Processing Technologies LLC

Log In
Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Visual Perception Processor and Method of Analyzing Parameters
  • Brief Description: The ’293 patent discloses a system and method for analyzing a visual scene or event by acquiring and processing one or more histograms of parameters (e.g., color, brightness, speed) associated with digitized picture elements, or "pixels." The system uses classifiers and validation logic to selectively include pixels in the histograms.

3. Grounds for Unpatentability

Ground 1: Claims 1, 18, 19, 22, and 29 are obvious over Pirim in view of Tomitaka.

  • Prior Art Relied Upon: Pirim (International Publication No. WO 99/36893) and Tomitaka (Patent 5,546,125).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Pirim, which discloses a system for detecting driver drowsiness using histograms, teaches the core architecture of the claimed processor. This includes a control unit, data busses, multiple histogram calculation units for different parameters (e.g., speed, direction), classifiers that compare pixel data to a criterion, and validation units that generate an enable signal for the histogram calculation. For limitations requiring classification based on statistical information from the histogram, Pirim discloses setting classification limits as a ratio of the histogram's maximum value. Tomitaka was cited to supplement this teaching, as it discloses a processor that calculates a classification threshold based on noise statistics derived from histogram outputs to validate which pixels are included in a hue histogram.
    • Motivation to Combine: Pirim explicitly suggests that its classification criteria can be set by a "separate computer program." Tomitaka, in the same technical field of histogram-based image processing, provides a working example of such a program. A POSITA would combine Tomitaka’s concrete implementation of a processor-calculated classification threshold to achieve the specific, stated goal in Pirim.
    • Expectation of Success: A POSITA would have a high expectation of success because Tomitaka provides a practical, successful implementation of a function that Pirim suggests in a highly analogous system.

Ground 2: Claims 1, 18, 19, 22, and 29 are obvious over Rogers in view of Gilbert.

  • Prior Art Relied Upon: Rogers (a 1978 Ph.D. dissertation on real-time video filtering) and Gilbert (a 1980 IEEE paper on a real-time video tracking system).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner asserted that Rogers' dissertation on missile tracking discloses a system that meets the claim limitations. Rogers’ system forms multiple histograms of a single parameter (pixel intensity) for different pre-defined regions of a tracking window (target, plume, background). It uses thresholds derived from these histograms to classify pixels as target, plume, or background. The results of this classification are then used to automatically update the position and size of the tracking window. These updated window definitions, in turn, serve as validation signals that determine which pixels are included in the next set of histograms. Gilbert was used to provide additional detail on the closed-loop feedback mechanism where a tracker processor controls the tracking window and feeds that data back to the video processor.
    • Motivation to Combine: Rogers and Gilbert describe the exact same missile tracking system, and Rogers is a co-author of the Gilbert paper. A POSITA studying the system in Rogers would be directly motivated to consult Gilbert for a more thorough description of the interaction between the system's video processor and tracker processor, which Gilbert provides.
    • Expectation of Success: Success would be expected because the references are not merely related but describe the same system from the same research group, ensuring their teachings are fully compatible and complementary.

Ground 3: Claims 1, 18, 19, 22, and 29 are obvious over Tomitaka in view of Rogers.

  • Prior Art Relied Upon: Tomitaka (Patent 5,546,125) and Rogers (a 1978 Ph.D. dissertation).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Tomitaka’s object tracking system discloses a processor with multiple histogram units that analyze different components of color (hue and brightness). Critically, Tomitaka validates the hue histogram based on a condition related to a different color property (saturation), preventing low-saturation pixels from being included. Rogers was introduced to teach the use of multiple histogram units for the same parameter (pixel intensity), validated based on a different pixel property (location within a tracking window). The combination teaches a system that uses multiple histogram units validated based on criteria different from the parameter being measured.
    • Motivation to Combine: A POSITA would be motivated to improve the capabilities of Tomitaka’s tracking system by incorporating the validation technique from Rogers. Specifically, a POSITA would apply Rogers' teaching of enabling a histogram based on a condition (like pixel location) that is different from the parameter being histogrammed (like color or intensity). This would enhance Tomitaka's ability to locate and track a moving object by adding another dimension of filtering.
    • Expectation of Success: A POSITA would expect success in combining these known techniques from the same field of histogram-based image tracking to enhance the functionality and robustness of the resulting system.

4. Relief Requested

  • Petitioner requests institution of an inter partes review and cancellation of claims 1, 18, 19, 22, and 29 of the ’293 patent as unpatentable.