Multiscale Contrast Enhancement Preserving Edge Transitions

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing multi-scale image processing methods for enhancing digital image contrast often distort grey value transitions, leading to unnatural appearances and artifacts, especially in medical images like CT scans, due to excessive non-linearity, resulting in overshoots at step edges and loss of homogeneity.

Innovation Solution

A new multi-scale contrast enhancement algorithm that computes translation difference images and adjusts pixel-wise amplification functions based on the ratio of enhanced to unenhanced center differences, applying these adjustments before summing translation difference images to preserve edge transitions and avoid distortions.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Illumination intensity

If conversion functions are applied to detail pixel values to enhance contrast, then image contrast is improved, but grey value transitions are distorted and artifacts are created

Engineering Contradiction:
Improveimage contrastVSAvoidgrey value transition accuracy
Core Design Contradiction:
Illumination intensityVSManufacturing precision

Solution Approach 1:

The patent segments the image processing into separate components: approximation images and detail images at multiple scales. The conversion function is applied selectively to detail pixel values rather than the entire image, allowing contrast enhancement while preserving overall structure. This segmentation enables localized modification of grey value transitions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent applies different processing quality to different regions: strong non-linear conversion is applied to detail pixel values where contrast enhancement is needed, while approximation images maintain their original linear characteristics. This local differentiation allows aggressive contrast enhancement in detail regions without distorting the global grey value transitions.

Inventive Principle:
Principle #3Local quality

2Illumination intensity

If non-linear conversion functions are applied to enhance contrast, then contrast enhancement is improved, but overshoots occur at step edges

Engineering Contradiction:
Improvecontrast enhancementVSAvoidedge transition shape
Core Design Contradiction:
Illumination intensityVSShape

Solution Approach 1:

The patent performs multi-scale decomposition before applying the non-linear conversion function. By separating the image into approximation and detail components at multiple scales, the system prepares the data structure that allows the conversion function to be applied in a controlled manner, preventing overshoots at edges while maintaining contrast enhancement.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent introduces the multi-scale dimension by decomposing the image into multiple resolution levels. This additional dimension allows the conversion function to operate on detail images at different scales independently, enabling contrast enhancement without propagating edge artifacts across the entire image.

Inventive Principle:
Principle #17Another dimension (Dimensionality change)

3Illumination intensity

If multi-scale techniques are applied to CT images with sharp grey level transitions, then contrast enhancement is improved, but artifacts become more significant

Engineering Contradiction:
Improvecontrast enhancementVSAvoidartifacts
Core Design Contradiction:
Illumination intensityVSObject-generated harmful factors

Solution Approach 1:

The patent segments CT images into approximation and detail components at multiple scales. This segmentation allows the non-linear conversion function to be applied selectively to detail pixel values, enhancing contrast in regions with sharp transitions while preserving the structural integrity that prevents artifact generation in the approximation regions.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes the processing parameters by applying different conversion functions at different scales. Strong non-linear conversion is applied to detail images where contrast enhancement is most needed, while approximation images maintain linear processing. This parameter differentiation reduces artifact generation in CT images with sharp grey level transitions.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP1933273B1Generating a contrast enhanced image using multiscale analysis
Publication Date: 2009.12.16 AGFA HEALTHCARE NV
  • EP1933273B1 patent drawingFigure 1
  • EP1933273B1 patent drawingFigure 2
  • EP1933273B1 patent drawingFigure 3

AI summary

A method of generating a multiscale contrast enhanced image is described wherein the shape of edge transitions is preserved. Detail images are subjected to a conversion, the conversion function of at least one scale being adjusted for each detail pixel value according to the ratio between the combination of the enhanced center differences and the combination of the unenhanced center differences.