Flash Non-Flash Image Segmentation Alignment Equalization

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing image acquisition systems face challenges in aligning and equalizing flash and non-flash images for improved image quality, as they often result in false contouring due to differences in pixel resolution and exposure levels, leading to suboptimal denoising, detail transfer, and red-eye correction.

Innovation Solution

A digital image acquisition system that uses a processor-controlled camera to capture and process flash and non-flash image pairs, employing segmentation filters, motion compensation, and alignment techniques to match pixel resolution and equalize exposure, thereby creating a higher quality synthesized image.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Reliability

If flash and non-flash images are directly combined for image processing, then denoising and detail transfer can be performed, but false contouring occurs due to differences in pixel resolution and exposure levels

Engineering Contradiction:
Improveimage qualityVSAvoidfalse contouring
Core Design Contradiction:
ReliabilityVSObject-affected harmful factors

Solution Approach 1:

The patent applies preliminary alignment and equalization actions before combining flash and non-flash images. The system performs alignment to match pixel resolutions and equalization to balance exposure levels between the two image sets, preventing false contouring artifacts from occurring during subsequent denoising and detail transfer operations

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system changes key parameters of the images including pixel resolution through alignment transformations and exposure levels through equalization processing. By adjusting these parameters to match between flash and non-flash images, the system enables reliable combination without introducing false contouring artifacts

Inventive Principle:
Principle #35Parameter changes

2Object-affected harmful factors

If pixel resolution matching is performed between flash and non-flash images, then false contouring is reduced, but processing complexity increases

Engineering Contradiction:
Improvefalse contouringVSAvoidprocessing complexity
Core Design Contradiction:
Object-affected harmful factorsVSDevice complexity

Solution Approach 1:

The patent segments the image processing task into distinct stages: alignment for resolution matching, equalization for exposure balancing, and final combination for denoising and detail transfer. This segmentation allows each stage to be optimized independently, reducing overall processing complexity while effectively eliminating false contouring

Inventive Principle:
Principle #1Segmentation

3Reliability

If exposure equalization is applied to flash and non-flash images, then image quality improves, but processing time increases

Engineering Contradiction:
Improveimage qualityVSAvoidprocessing time
Core Design Contradiction:
ReliabilityVSLoss of time

Solution Approach 1:

The system performs exposure equalization as a preliminary step before the computationally intensive denoising and detail transfer operations. By equalizing exposure levels early, the system ensures optimal input quality for subsequent processing stages while minimizing total processing time through efficient algorithm design

Inventive Principle:
Principle #10Preliminary action

Data Source

PatentEP1918872B1Image segmentation method and system
Publication Date: 2009.07.22 FOTONATION VISION
  • EP1918872B1 patent drawingFigure 1
  • EP1918872B1 patent drawingFigure 2(a)
  • EP1918872B1 patent drawingFigure 2(b)

AI summary

A digital segmentation method and apparatus determines foreground and/or background within at least one portion of a captured image. The determining includes comparing a captured image to a pre-captured or post captured reference image of nominally the same scene. One of the images is taken with flash and the other without. The system can be implemented as part of a digital camera acquisition chain having effective computation complexity,