Multispectral Illumination Control for Real-Time Contrast Imaging

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional imaging systems fail to detect individual spectral components of incident light, leading to inadequate scene-specific or scene-general contrast for distinguishing objects that appear similar in nature.

Innovation Solution

An imaging system adjusts the intensities of each spectral band of illumination based on a classification vector that separates reflectance values between objects, generating an image with optimized contrast by tuning the illumination intensities to enhance the discernibility of different materials.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Ease of operation

If conventional imaging processes integrate over spectral dimensions, then imaging simplicity is maintained, but contrast for distinguishing objects is lost

Engineering Contradiction:
Improveimaging simplicityVSAvoidcontrast discrimination
Core Design Contradiction:
Ease of operationVSMeasurement precision

Solution Approach 1:

The patent segments the integrated spectral information into discrete wavelength components by using multiple illumination sources at different wavelengths. Instead of physically integrating over the spectrum, the system captures separate images at discrete wavelengths and processes them computationally to achieve contrast enhancement while maintaining operational simplicity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent introduces classification vectors as an intermediary between the raw spectral data and the final image. These vectors serve as mediators that encode spectral signatures of different objects, allowing the system to enhance contrast by projecting pixel intensities onto these vectors without requiring complex spectral integration hardware.

Inventive Principle:
Principle #24Intermediary (Mediator)

2Measurement precision

If multiple wavelength illumination is used, then spectral discrimination capability is improved, but system complexity increases

Engineering Contradiction:
Improvespectral discriminationVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent employs tunable LED illumination sources that can dynamically adjust their emission wavelengths and intensities. This dynamic capability allows the system to selectively illuminate at specific wavelengths of interest and adapt the illumination profile based on the classification vectors, achieving spectral discrimination without requiring a fixed complex multi-wavelength system.

Inventive Principle:
Principle #15Dynamics

Solution Approach 2:

The system changes the parameters of the illumination sources (wavelength and intensity) based on the classification vectors. By adjusting these parameters dynamically, the system can emphasize spectral regions where different objects have distinct reflectance characteristics, achieving high spectral discrimination with a relatively simple illumination system.

Inventive Principle:
Principle #35Parameter changes

3Ease of manufacture

If pixel intensity is determined by physical integration, then imaging process is straightforward, but scene-specific contrast is lost

Engineering Contradiction:
Improveimaging process simplicityVSAvoidscene-specific contrast
Core Design Contradiction:
Ease of manufactureVSLoss of information

Solution Approach 1:

The patent performs preliminary action by pre-computing classification vectors from training data that capture the spectral signatures of different objects in the scene. These pre-computed vectors are then used to guide the illumination and image processing, allowing the system to preserve and enhance scene-specific contrast without requiring complex real-time spectral analysis.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The system uses classification vectors as feedback to adjust the illumination intensities at different wavelengths. By projecting pixel intensities onto these vectors and using the results to guide further illumination adjustments, the system creates a feedback loop that preserves and enhances scene-specific contrast information that would otherwise be lost in conventional integration.

Inventive Principle:
Principle #23Feedback

Applied Scientific Principles

This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.

Function Achieved in This Case

The system produces images with significantly enhanced contrast between objects, allowing clear differentiation of anatomical structures or materials that appear similar under conventional illumination, improving surgical and medical imaging accuracy.

Implementation Method 1

one or more LED sources of the plurality of spectral illumination sources may be illuminated with a combination of light from two or more spectral bands

Methodology Applied
Scientific EffectLight emission from LED sources: Light Emitting Diode

Implementation Method 2

separating a plurality of collected reflectance values for the first object from a plurality of collected reflectance values for the second object

Methodology Applied
Scientific EffectLight reflection: Reflection

Data Source

PatentEP4535292B1Optical implementation of machine learning for real time increased contrast via multiple wavelength illumination with tunable power
Publication Date: 2026.04.29 VERILY HEALTH INC
  • EP4535292B1 patent drawingFigure 1
  • EP4535292B1 patent drawingFigure 2A~2B
  • EP4535292B1 patent drawingFigure 3

AI summary

An imaging system (e.g., hyperspectral imaging system) receives an indication to compare a first object and a second object (e.g., two anatomical structures or organs in a medical environment). The imaging system accesses a classification vector for the first object and the second object, the classification vector having been extracted by separating a plurality of collected reflectance values for the first object from a plurality of collected reflectance values for the second object. A set of optimal illumination intensities for one or more spectral illumination sources of the imaging system is determined based on the extracted classification vector. The first and second objects are illuminated with the determined illumination intensities. A high-contrast image of the first and second objects is provided for display, such that the two objects can be readily distinguished in the image. The intensity of pixels in the image is determined by the illumination intensities.