Super-Resolution Image Generation Using Optical Flow Segmentation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Existing methods for generating super-resolution (SR) images in non-visible spectral ranges face challenges in increasing image resolution without enhancing detector resolution, particularly when dealing with moving or changing objects, which can result in artifacts and require discarding image sequences.

Innovation Solution

The method calculates optical flow for individual images, segments them based on displacement vector fields, and optimizes the SR image using separate variation parameters for each segment, allowing adaptation to varying image content and improving robustness by masking irrelevant sections.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If the resolution of the detector is increased to improve image resolution, then the image resolution is improved, but the production costs increase significantly

Engineering Contradiction:
Improveimage resolutionVSAvoidproduction costs
Core Design Contradiction:
Measurement precisionVSEase of manufacture

Solution Approach 1:

The patent divides the image into multiple segments or regions, and applies different processing strategies to different segments. This allows the system to focus computational resources on critical areas while using simpler methods for other areas, achieving high resolution without requiring a high-resolution detector across the entire image.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent changes parameters such as exposure time, illumination intensity, or processing algorithms to achieve better image quality without improving detector resolution. By adjusting these parameters, the system can extract more information from the same detector hardware.

Inventive Principle:
Principle #35Parameter changes

2Measurement precision

If small deviations between individual images are used to generate SR images, then higher image resolution is achieved, but artifacts appear when objects move or change significantly

Engineering Contradiction:
Improveimage resolutionVSAvoidrobustness to object changes
Core Design Contradiction:
Measurement precisionVSReliability

Solution Approach 1:

The patent applies different processing qualities or methods to different regions of the image based on local characteristics. For regions with significant object changes, more robust methods are applied, while for static regions, standard SR techniques are used. This local adaptation maintains reliability across varying conditions.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The patent introduces dynamic adaptation in the SR generation process, where the processing parameters and methods are adjusted based on the detected motion or changes in the scene. This allows the system to respond to object movements and changes, maintaining reliability while achieving high resolution.

Inventive Principle:
Principle #15Dynamics

3Device complexity

If a single optimization method is applied to all image areas, then the process is simple, but image areas with high variability are not processed optimally

Engineering Contradiction:
Improveprocessing complexityVSAvoidprocessing quality
Core Design Contradiction:
Device complexityVSMeasurement precision

Solution Approach 1:

The patent segments the image into multiple regions with different characteristics, and applies tailored optimization methods to each segment. This segmentation approach balances complexity and quality by focusing advanced processing only where needed.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The patent implements local quality optimization by adapting the processing method to the specific characteristics of each image region. This allows high-quality processing for variable regions while maintaining simplicity for uniform regions, achieving an optimal balance between complexity and precision.

Inventive Principle:
Principle #3Local quality

Data Source

PatentEP2791896B1Method for generating super-resolution images having improved image resolution and measuring device
Publication Date: 2015.10.21 TESTO AG
  • EP2791896B1 patent drawingFigure 1
  • EP2791896B1 patent drawingFigure 2~3

AI summary

In the case of a measuring device for recording a sequence of individual images (6, 7, 8) in a non-visible spectral range, a method for generating an SR image (11) having an image resolution that is higher than an image resolution of the individual images (6, 7, 8) is proposed, wherein, for the individual images (6, 7, 8), a displacement vector field (11, 12) is determined with a calculation of the optical flow and the individual images (6, 7, 8) are segmented into segments (20, 21, 22, 23, 24, 25) with regard to the values of the displacement vector field (11, 12), wherein an optimization method is carried out for calculating the SR image (11) from the individual images (6, 7, 8) with variation parameters individually assigned to the segments (20, 21, 22, 23, 24, 25).