PTAB

IPR2018-01802

Samsung Electronics Co Ltd v. FotoNation Ltd

Log In
Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Digital Image Processing and Enhancement
  • Brief Description: The ’715 patent relates to digital image processing, specifically methods for enhancing auto-focus in a digital acquisition device. The technology involves using information gained from detecting and tracking a region of interest, such as a face, within a stream of digital images to automatically adjust the focus position.

3. Grounds for Unpatentability

Ground 1: Anticipation by Hashimoto - Claims 1 and 6 are anticipated by Hashimoto under 35 U.S.C. §102.

  • Prior Art Relied Upon: Hashimoto (Patent 6,072,526).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Hashimoto, which was not considered during prosecution, disclosed every limitation of claims 1 and 6. Hashimoto described a video camera that enhances auto-focus by detecting and tracking a person's face. It identified a skin-colored portion as a region of interest, tracked its movement across successive video frames, and automatically adjusted the auto-focus (AF) area and lens position based on this tracking. Petitioner asserted this process met the claim limitations of identifying a region of interest, determining its degree of focus, tracking it within a stream of images, detecting that its focus position has changed based on movement, and automatically adjusting the focus position accordingly. For claim 6, Petitioner argued Hashimoto estimated an auto-focus distance by determining the size of the AF area needed to cover the detected face, thereby linking the size of the region of interest to the focus distance.

Ground 2: Obviousness over Hashimoto and Chen - Claims 2-4 are obvious over Hashimoto in view of Chen under 35 U.S.C. §103.

  • Prior Art Relied Upon: Hashimoto (Patent 6,072,526) and Chen (Patent 7,508,961).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner contended that Hashimoto taught the base method of claim 1, including identifying one or more candidate regions of interest (e.g., multiple faces). Chen taught a method for improving face detection accuracy by "weighted averaging of the locations and scales of all tests that indicated face presence." Petitioner argued that this weighted averaging corresponded to the limitation in claim 2 of computing a weighted average of candidate regions. For claims 3 and 4, Chen's teaching of weighting by "locations and scales" (size) and "classifier scores" (likelihood of a face) directly mapped to assigning an "estimated importance" and a "relative value of detection assurance."
    • Motivation to Combine: A POSITA would combine Chen's weighted averaging technique with Hashimoto's face-tracking auto-focus system to solve a known problem: improving face detection speed and accuracy, particularly when multiple faces are present. This would have been a predictable application of a known image processing technique to improve a known system.
    • Expectation of Success: A POSITA would have had a reasonable expectation of success because combining these known face detection functionalities was a mere application of known technologies to produce a predictable result (improved detection accuracy in Hashimoto's auto-focus process).

Ground 3: Anticipation by Yamashita - Claim 1 is anticipated by Yamashita under 35 U.S.C. §102.

  • Prior Art Relied Upon: Yamashita (Patent 7,423,686).

  • Core Argument for this Ground:

    • Prior Art Mapping: Petitioner argued that Yamashita, also not considered during prosecution, disclosed an "auto-focus method for estimating and tracking the motion of a region of interest (ROI) while continuously focusing the object automatically." Yamashita's system specified an ROI, determined its position and distance over initial frames, and then used a feedback loop to estimate the ROI's next location and distance based on past and present information. This estimated future position was then used to control the camera's focus. Petitioner asserted this directly taught the claimed steps of identifying a group of pixels (the ROI), determining a degree of focus (position and distance), tracking the ROI in a stream of images, detecting movement (by calculating an error between estimated and actual positions), and automatically adjusting focus based on the estimated values.

  • Additional Grounds: Petitioner asserted additional obviousness challenges based on Hashimoto in view of Nakamura (Japanese Patent Publication 2002-296489) and Ide (Patent 6,785,469), and based on Yamashita in view of Chen, Nakamura, and Ide. These grounds relied on similar arguments for combining known image processing techniques to add functionalities like weighted averaging or distance estimation based on object size.

4. Relief Requested

  • Petitioner requested institution of an inter partes review and cancellation of claims 1-4 and 6 of the ’715 patent as unpatentable.