PTAB

IPR2025-01519

Samsung Electronics America Inc v. SnapAid Ltd

Log In
Key Events
Petition
petition

1. Case Identification

2. Patent Overview

  • Title: Real Time Assessment of Picture Quality
  • Brief Description: The ’901 patent describes a method and system for real-time assessment of image quality in camera devices. The system uses multiple sensors to determine a plurality of quality indicators (QIs) for an image, combines these indicators to evaluate the total image quality, and provides real-time suggestions to the user for improvement.

3. Grounds for Unpatentability

Ground 1: Claims 1-4, 6, and 8 are obvious over Anon, Suzuki, Jasinski, and Garcia-Molina.

  • Prior Art Relied Upon: Anon (Patent 8,508,622), Suzuki (Patent 5,831,670), Jasinski (Application # 2012/0201427), and Garcia-Molina (a 2009 textbook on database systems).
  • Core Argument for this Ground:
    • Prior Art Mapping: Petitioner argued that Anon taught the foundational system of independent claim 1: a camera device with sensors that estimates image quality by computing a "composition measure" from multiple image-based characteristics and provides real-time feedback. To the extent Anon did not explicitly teach obtaining a value for the device's angle to the horizon, Suzuki remedied this by disclosing a camera that detects tilt and provides corrective feedback. For claim limitations requiring the dynamic "estimating" of weights, Jasinski was cited for its teaching of dynamically computing weights for image regions based on current scene conditions. Finally, Garcia-Molina was used to show the obviousness of storing suggestions in a "pre-stored table," as it taught that tables are a common and fundamental data structure.
    • Motivation to Combine: A POSA would combine these references to improve the functionality of Anon’s system. Suzuki’s well-known tilt-correction feature would be a natural addition to Anon's composition analysis to improve aesthetic quality. Jasinski’s dynamic weighting would be incorporated to make Anon’s quality assessment more accurate and adaptive to changing scene conditions, a predictable improvement. Implementing Anon’s "rule sets" as tables, as suggested by Garcia-Molina, represented a simple design choice.
    • Expectation of Success: Petitioner asserted a POSA would have a high expectation of success, as the combination involved integrating known, modular techniques (e.g., tilt correction, dynamic weighting) into a system designed to be extensible, yielding predictable improvements in image quality assessment.

Ground 2: Claims 10, 13-14, and 17-20 are obvious over Anon, Takeuchi, Jasinski, Kosaka, and Garcia-Molina.

  • Prior Art Relied Upon: Anon (Patent 8,508,622), Takeuchi (Application # 2010/0149361), Jasinski (Application # 2012/0201427), Kosaka (Application # 2004/0012682), and Garcia-Molina (a 2009 textbook).
  • Core Argument for this Ground:
    • Prior Art Mapping: This ground targeted independent claim 10, which recites a method for estimating quality from a video stream. Anon was again asserted as the base system, as it disclosed its methods could apply to videography. Takeuchi was introduced to teach obtaining a value for under- or over-exposure, a fundamental quality metric not expressly detailed in Anon. Kosaka was cited for its disclosure of a system that detects lens obstruction (e.g., a finger), mapping to the claim limitation of obtaining a value responsive to lens obstruction. Jasinski and Garcia-Molina served the same roles as in Ground 1, providing dynamic weighting and the obviousness of using tables, respectively.
    • Motivation to Combine: A POSA would be motivated to create a more comprehensive and robust quality assessment tool by incorporating additional, fundamental quality metrics into Anon's framework. Adding Takeuchi’s exposure analysis and Kosaka’s obstruction detection would fill clear gaps in Anon’s system, leading to a more complete evaluation of image quality. Both are common image flaws, and their detection would be a logical enhancement.
    • Expectation of Success: The combination was presented as a straightforward integration of distinct, modular image analysis functions into an extensible platform. A POSA would expect success in adding these well-understood quality checks to produce a more powerful and commercially desirable product.

Ground 3: Claims 5 and 7 are obvious over Anon, Suzuki, Jasinski, Garcia-Molina, and Ramesh.

  • Prior Art Relied Upon: The combination from Ground 1, further in view of Ramesh (Application # 2009/0296989).
  • Core Argument for this Ground:
    • Prior Art Mapping: This ground built upon the combination for claim 1 by adding Ramesh to address the limitations of dependent claims 5 and 7. These claims require associating a weight with an "estimated error" and using a "probability distribution function." Ramesh taught a computer vision system that uses a probabilistic framework to track objects, explicitly teaching that "less certain guesses" (i.e., those with greater error) are "weighted less."
    • Motivation to Combine: Petitioner argued a POSA would be motivated to incorporate Ramesh's probabilistic approach to improve the accuracy and robustness of Anon's weighting system. By associating weights with an estimated error, Anon's system could more intelligently handle uncertainty from noisy or ambiguous sensor data, a known technique for improving such evaluation engines.
    • Expectation of Success: Given that Anon's system was designed to be flexible and improved over time, incorporating a known statistical technique from Ramesh to handle uncertainty would be a predictable path to enhancement with a high expectation of success.
  • Additional Grounds: Petitioner asserted additional obviousness challenges based on combinations including Liu (an IEEE 2012 publication) for its teachings on fusing multiple aesthetic algorithms like PSNR, and Alhadef (Patent 8,009,198) for its teachings on using Kalman filters to compensate for sensor drift and error when weighting sensor inputs.

4. Relief Requested

  • Petitioner requests institution of IPR and cancellation of claims 1-20 of the ’901 patent as unpatentable.