Super-Resolved Depth Mapping From Low-Resolution Image Frames

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Aerial images and videos captured by low-resolution image sensors often result in poor depth information and 3D representations due to their low image quality, which is inadequate for applications requiring high-resolution images, such as computer vision and mixed-reality experiences.

Innovation Solution

A system that groups low-resolution image frames based on pose similarity and applies super-resolution processing to generate high-resolution composite images, which are then used for depth processing techniques like stereo matching to produce accurate depth maps.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Quantity of substance

If low-resolution image sensors are used for aerial imaging, then storage and bandwidth requirements are reduced, but depth information quality deteriorates

Engineering Contradiction:
Improvestorage and bandwidth requirementsVSAvoiddepth information quality
Core Design Contradiction:
Quantity of substanceVSMeasurement precision

Solution Approach 1:

The patent combines multiple low-resolution images into a single high-resolution composite image by merging corresponding pixels from multiple input images. This merging process reconstructs fine details that were lost in individual low-resolution images, enabling high-quality depth information extraction while maintaining the advantage of using low-resolution sensors for reduced storage and bandwidth requirements.

Inventive Principle:
Principle #5Merging (Combining)

2Measurement precision

If high-resolution image sensors are used for aerial imaging, then depth information quality is improved, but storage and bandwidth requirements increase

Engineering Contradiction:
Improvedepth information qualityVSAvoidstorage and bandwidth requirements
Core Design Contradiction:
Measurement precisionVSQuantity of substance

Solution Approach 1:

The patent creates a high-resolution copy of the scene by synthesizing a composite image from multiple low-resolution images. Instead of capturing high-resolution images directly (which would require high-resolution sensors and generate large data volumes), the system copies and processes multiple low-resolution images to reconstruct high-resolution detail, thereby achieving high depth information quality without the storage and bandwidth burden of actual high-resolution captures.

Inventive Principle:
Principle #26Copying

3Measurement precision

If multiple low-resolution images are processed to generate high-resolution composite images, then depth information quality is improved, but processing complexity increases

Engineering Contradiction:
Improvedepth information qualityVSAvoidprocessing complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent segments the image processing task into distinct stages: capturing multiple low-resolution images, aligning them based on pose information, combining corresponding pixels to generate the composite image, and finally extracting depth information. This segmentation of the processing workflow manages complexity by breaking down the overall task into manageable, sequential operations rather than attempting to perform all processing in a single complex step.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP4323964B1Systems and methods for generating depth information from low-resolution images
Publication Date: 2025.12.31 MICROSOFT TECHNOLOGY LICENSING LLC
  • EP4323964B1 patent drawingFigure 1
  • EP4323964B1 patent drawingFigure 2A
  • EP4323964B1 patent drawingFigure 2B

AI summary

A system for generating depth information from low-resolution images is configured to access a plurality of image frames capturing an environment, identify a first group of image frames from the plurality of image frames, and generate a first image comprising a first composite image of the environment using the first group of image frames as input. The first composite image has an image resolution that is higher than an image resolution of the image frames of the first group of image frames. The system is also configured to obtain a second image of the environment, where parallax exists between a capture perspective associated with the first image and a capture perspective associated with the second image. The system is also configured to generate depth information for the environment based on the first image and the second image.