Light Source Estimation via Iterative Rendering and Error Minimization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for estimating light sources in real scenes are inefficient, require significant user effort, and often result in unstable processing due to high computational load and local optimization issues, especially when comparing computer-generated and actual images.

Innovation Solution

A camera-based information processing device that estimates light sources by acquiring image data, shape data, and image capturing conditions, using an initial light source data determination and derivation process to optimize light source data through iterative rendering and error minimization, allowing for efficient and stable light source estimation without the need for specialized apparatus or extensive user movement.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the method by Japanese Patent Laid-Open No. 2008-33531 is used to estimate light source by acquiring entire periphery image, then light source estimation can be performed, but the user must move in the entire periphery direction which requires significant time and effort

Engineering Contradiction:
Improvelight source estimation accuracyVSAvoidtime required for image acquisition
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent pre-calculates and stores light source distribution data for various camera positions and orientations before actual image capture. This preliminary preparation allows the system to quickly retrieve and compare pre-rendered images with the captured image, eliminating the need for time-consuming real-time rendering or user movement to acquire multiple images.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent creates computer-generated copy images based on the captured image and compares them to estimate the light source. By generating synthetic images with controlled lighting conditions and comparing them to the actual captured image, the system can determine the most likely light source configuration without requiring physical movement or additional image capture.

Inventive Principle:
Principle #26Copying

2Measurement precision

If the method by Okabe is used to estimate light source by comparing CG image data with actually photographed image data, then light source parameters can be optimized, but the large number of light source parameters causes optimization to take much time and results in local solutions making processing unstable

Engineering Contradiction:
Improvelight source parameter accuracyVSAvoidprocessing stability
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent extracts only the essential light source parameters needed for accurate estimation, removing unnecessary parameters from the optimization process. By focusing on key parameters such as light source position, intensity, and color temperature while excluding redundant parameters, the system achieves stable and reliable results without falling into local optimization traps.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent transforms the light source parameter representation to a more manageable form that reduces the search space for optimization. By changing how light source parameters are defined and represented, the system can efficiently find the global optimum without being constrained by the large parameter space that causes instability in traditional methods.

Inventive Principle:
Principle #35Parameter changes

3Manufacturing precision

If traditional image processing is performed to change light source kind and direction or synthesize virtual objects, then the intended effect can be achieved, but a skilled person must spend much time to create a processed image without unnaturalness

Engineering Contradiction:
Improveimage processing qualityVSAvoidimage processing speed
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The patent enables automatic light source estimation and image processing without requiring skilled manual intervention. The system autonomously analyzes the captured image, estimates light source parameters, generates appropriate processed images, and ensures naturalness through automated comparison with pre-calculated reference data, thereby achieving high quality results rapidly.

Inventive Principle:
Principle #25Self-service

Solution Approach 2:

The patent employs feedback mechanisms where the system continuously compares generated processed images with the original captured image and pre-calculated reference images to ensure naturalness. This automated feedback loop allows the system to adjust processing parameters in real-time, achieving high-quality results without manual intervention and significantly improving processing speed.

Inventive Principle:
Principle #23Feedback

Data Source

PatentEP2662833B1Light source data processing device, method and program
Publication Date: 2021.01.06 CANON KK
  • EP2662833B1 patent drawingFigure 1
  • EP2662833B1 patent drawingFigure 2
  • EP2662833B1 patent drawingFigure 3

AI summary

Conventional calculation of light source data requires a high load and is unstable. An information processing device derives light source data which represents a state of a light source of an image represented by a captured image based on image capturing condition data at the time of capturing the image of a subject.