PTAB

IPR2018-01799

Samsung Electronics Co., Ltd. v. FotoNation Limited

Log In

1. Case Identification

  • Patent #: 7,574,016
  • Filed: September 27, 2018
  • Petitioner(s): Samsung Electronics Co., Ltd.
  • Patent Owner(s): FotoNation Limited
  • Challenged Claims: 1 and 2

2. Patent Overview

  • Title: Digital Image Processing With Face Detection
  • Brief Description: The ’016 patent relates to digital image processing, specifically methods and systems for automatically processing or enhancing a digital image by using information gained from identifying and analyzing one or more faces within the image. The technology involves using face detection to adjust image parameters like focus, color, and exposure.

3. Grounds for Unpatentability

Ground 1: Claims 1 and 2 are unpatentable under 35 U.S.C. §102(e) as anticipated by Hara.

  • Prior Art Relied Upon: Hara (Patent 7,170,633).
  • Core Argument:
    • Prior Art Mapping: Petitioner argued that Hara teaches every limitation of claims 1 and 2. Hara describes a portable digital camera that processes images using face detection to achieve desired image parameters. Petitioner contended Hara’s system acquires a “temporally-sequential collection” of images by continuously capturing and processing images in a pre-shooting, viewfinder-like mode. Hara’s “scene judger” identifies a face, and the system tracks its movement by comparing subsequent frames. Hara determines “default values” by establishing predefined parameters for color balance and luminance and modifies a current image’s parameters (e.g., exposure value, color) based on an analysis of the detected face region against these defaults. The modified image is then displayed. For claim 2, Petitioner asserted that Hara’s “thinning operation,” which reduces the number of pixels for processing, constitutes a “collection of images” that have a lower resolution than the main acquired image.

Ground 2: Claim 1 is unpatentable under 35 U.S.C. §102(b) as anticipated by Hashimoto.

  • Prior Art Relied Upon: Hashimoto (Patent 6,072,526).
  • Core Argument:
    • Prior Art Mapping: Petitioner contended that Hashimoto, which describes a portable video camera, discloses all elements of claim 1. Hashimoto’s camera performs a “person detection” routine to identify a face by analyzing picture elements for skin color and a circular shape. It acquires a “temporally-sequential collection” of images as it processes a continuous video signal. The system identifies and tracks a face across this collection by detecting a skin-colored portion in a new frame near the auto-focus (AF) area of a previous frame and moving the AF area accordingly. Hashimoto determines “default values” by using a predefined target luminance range (e.g., 55IRE to 75IRE) for skin tones. It then modifies parameters in the main acquired image by adjusting the camera’s iris or automatic gain control (AGC) to bring the detected face’s luminance within that default range, thereby adjusting exposure. Finally, Hashimoto discloses storing the modified video signal on a cassette deck.

Ground 3: Claim 2 is obvious over Hashimoto in view of Edanami.

  • Prior Art Relied Upon: Hashimoto (Patent 6,072,526) and Edanami (Patent 6,297,846).
  • Core Argument:
    • Prior Art Mapping: Petitioner argued that while Hashimoto teaches the method of claim 1, it does not explicitly disclose using a collection of lower resolution images for face detection, as required by claim 2. Edanami, which relates to videoconferencing systems, was argued to supply this missing element. Edanami explicitly teaches capturing video, storing it in a frame memory, and then creating a lower-resolution, “subsampled” version of the video in a separate memory. This subsampled, lower-resolution version is then used for face position detection, while the full-resolution image is used for final display or storage. Edanami explains this is done to "reduce the computation time" for the face detection process.
    • Motivation to Combine: A person of ordinary skill in the art (POSITA) would combine Edanami's well-known technique for improving computational efficiency with Hashimoto's face detection and image correction system. Since both references operate in the same field of face detection and tracking in video, a POSITA would have been motivated to incorporate Edanami's subsampling method into Hashimoto's system to speed up the face detection process without sacrificing the quality of the final, recorded image.
    • Expectation of Success: The combination would have yielded the predictable result of a more computationally efficient face detection and tracking system. Applying a known technique (subsampling to reduce processing load) to a similar system (Hashimoto's camera) to achieve its stated benefit (reduced computation time) would have been a straightforward modification with a high expectation of success.

4. Relief Requested

  • Petitioner requests the institution of an inter partes review and the cancellation of claims 1 and 2 of the ’016 patent as unpatentable.