Plenoptic Multifocal Eye Image Fusion via Gradient Analysis

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current eye imaging technologies face challenges in capturing well-focused images throughout the thickness of the retina or ocular regions due to optical aberrations and alignment issues, resulting in low image quality and difficulty in combining images from different focal planes.

Innovation Solution

A method that combines multiple eye images into a multifocal image using image processing algorithms to register and align images, identify in-focus regions, and select corresponding intensities, improving resolution and image quality by creating a plenoptic multifocal image that can be used with various eye imaging modalities.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If traditional single-focus imaging is used, then the imaging process is simple, but image quality throughout the thickness of the retina is poor

Engineering Contradiction:
Improveimage qualityVSAvoidimaging process
Core Design Contradiction:
Manufacturing precisionVSDevice complexity

Solution Approach 1:

The patent divides the imaging process into multiple focal plane acquisitions, segmenting the retinal thickness into discrete focal layers. Each layer is captured separately with optimized focus, then computationally recombined to form a comprehensive multifocal image that preserves detail across the entire retinal thickness.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent transitions from two-dimensional single-plane imaging to three-dimensional multifocal imaging by adding the depth dimension. Multiple images at different focal planes are acquired and processed to create a plenoptic representation that encodes depth information, enabling visualization of structures at various retinal depths.

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

2Manufacturing precision

If multiple images at different focal planes are acquired, then overall image quality improves, but alignment and registration become difficult

Engineering Contradiction:
Improveimage qualityVSAvoidalignment accuracy
Core Design Contradiction:
Manufacturing precisionVSMeasurement precision

Solution Approach 1:

The patent employs feedback mechanisms through image registration algorithms that automatically adjust alignment based on detected features across multiple focal planes. The system iteratively refines the positioning and orientation of each captured image to achieve precise registration, using feedback from edge detection and feature matching to correct misalignments.

Inventive Principle:
Principle #23Feedback

Solution Approach 2:

The patent introduces computational image processing algorithms as an intermediary between image acquisition and final visualization. These algorithms perform automatic registration, warping correction, and fusion operations, acting as a mediator that reconciles the geometric discrepancies between images taken at different focal planes and angles.

Inventive Principle:
Principle #24Intermediary (Mediator)

3Manufacturing precision

If focus control is increased to capture multiple focal planes, then depth coverage improves, but optical aberrations increase

Engineering Contradiction:
Improvedepth coverageVSAvoidoptical aberrations
Core Design Contradiction:
Manufacturing precisionVSObject-affected harmful factors

Solution Approach 1:

The patent extracts and removes the harmful effects of optical aberrations through computational processing. By separating the in-focus regions from out-of-focus regions using gradient analysis and frequency domain filtering, the system extracts only the useful in-focus information while discarding the blurred, aberrated portions of each image.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent changes the parameter of focus across multiple acquisitions, capturing images at systematically varied focal depths. By varying the focus parameter and then computationally recombining these images, the system achieves comprehensive depth coverage while using image processing to correct the aberrations inherent in each individual focal plane capture.

Inventive Principle:
Principle #35Parameter changes

Data Source

PatentEP2787877B1Method for combining a plurality of eye images into a plenoptic multifocal image
Publication Date: 2022.04.06 NEUROVISION IMAGING LLC
  • EP2787877B1 patent drawingFigure 1
  • EP2787877B1 patent drawingFigure 2
  • EP2787877B1 patent drawingFigure 3

AI summary

A method for combining a plurality of eye images into a plenoptic multifocal image that includes registering the eye images with a plurality of frames into one or more eye image sets with a processor and a memory system, aligning each of the eye images in each of the one or more image sets with a selected reference that resides on the memory system with the processor and determining one or more in-focus regions of the eye images by calculating one or more gradient images while ignoring noise and other imaging artifacts. The method also includes identifying the one or more in-focus regions with highest resolution from the one or more gradient images and selecting one or more corresponding in-focus intensities from the frames to combine into a plenoptic multifocal image with a higher resolution than the eye images, the frames and the one or more eye image sets.