Endoscope Super-Resolution Using Simulated Optical Imaging

Resolve Bottlenecks,
Find Innovative SolutionsGenerate Solutions

Solution Overview

Problem

Existing super resolution techniques face challenges in accurately recovering high resolution images from low resolution images, especially in medical endoscopes where the small imager size results in low resolution, and the imaging systems of high and low resolution sensors differ significantly.

Innovation Solution

The proposed information processing system uses a trained model to recover the resolution of images captured through a second imaging system with fewer pixels than a first imaging system. This is achieved by simulating the imaging method and optical system of the second imaging system during the training process, allowing the model to accurately recover the high resolution image.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Length of moving object

If a small image sensor is used in an endoscope to reduce probe diameter, then the endoscope can access narrower body cavities, but the image resolution deteriorates

Engineering Contradiction:

Improveprobe diameterVSAvoidimage resolution

Core Design Contradiction:

Length of moving objectVSMeasurement precision

Solution Approach 1:

The patent creates a trained model that copies the imaging characteristics of a high-resolution first imaging system into a low-resolution second imaging system. By simulating the optical system and imaging method during training, the model learns to map low-resolution images to high-resolution images, effectively copying the quality characteristics without physically copying the hardware.

Inventive Principle:

Principle #26Copying

Solution Approach 2:

The patent changes the resolution parameter of the image by using a trained model that transforms low-resolution images to high-resolution images. The model adjusts the resolution parameter through deep learning, converting images from the state captured by the small sensor to the equivalent quality of a larger sensor system.

Inventive Principle:

Principle #35Parameter changes

2Ease of manufacture

If deep learning is performed using simply reduced low resolution images, then the training process is simple, but the resolution recovery accuracy deteriorates because the imaging system differences are not considered

Engineering Contradiction:

Improvetraining process complexityVSAvoidresolution recovery accuracy

Core Design Contradiction:

Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent performs preliminary simulation of the imaging method and optical system during the training stage. By pre-simulating how the second imaging system would capture images and how the first imaging system would capture the corresponding high-resolution images, the training model is prepared in advance with accurate imaging characteristics, leading to better resolution recovery accuracy.

Inventive Principle:

Principle #10Preliminary action

Solution Approach 2:

The patent introduces an intermediary simulation process that mediates between the low-resolution input and high-resolution output. The simulation acts as an intermediary layer that translates the imaging characteristics of the second system into the reference framework of the first system, enabling accurate training without directly comparing dissimilar imaging outputs.

Inventive Principle:

Principle #24Intermediary (Mediator)

3Measurement precision

If optical system simulation processing is added to simulate the second imaging method during training, then the resolution recovery accuracy improves, but the training process complexity increases

Engineering Contradiction:

Improveresolution recovery accuracyVSAvoidtraining process complexity

Core Design Contradiction:

Measurement precisionVSDevice complexity

Solution Approach 1:

The patent replaces complex physical optical systems with computational simulation during training. Instead of using multiple physical imaging systems with different optical characteristics, the patent uses software-based simulation to model the optical system behavior, substituting mechanical/optical complexity with computational processing.

Inventive Principle:

Principle #28Mechanics substitution (Replace mechanical system)

Data Source

PatentUS12347066B2Information processing system, endoscope system, and information storage medium

Publication Date: 2025.07.01 OLYMPUS CORPORATION(JP)

AI summary

An information processing system includes a processor. The trained model is trained to resolution recover a low resolution training image generated by low resolution processing performed on a high resolution training image to a high resolution training image that represents a high resolution image captured with a predetermined object through the first imaging system. The low resolution processing represents processing that generates a low resolution image as if captured with the predetermined object through the second imaging system and processing that simulates the second imaging method, and includes processing that simulates a resolution characteristic of an optical system of the second imaging system. The processor uses the trained model to resolution recover the processing target image captured through a second imaging system to an image having a resolution at which the first imaging system performs imaging.