3D Coordinate Determination Using Time-Dependent Bundle Adjustment

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for determining three-dimensional coordinates from two-dimensional images lack the ability to accurately account for time-dependent changes in a scene, such as vegetation growth or geological movements, leading to incomplete and inaccurate height determinations.

Innovation Solution

A computer-implemented method that uses bundle-adjustments based on pair-wise measurements of images from multiple viewpoints, incorporating time as a parameter to minimize re-projection error and allow for locally time-dependent height variations, thereby refining 3D point positions and camera parameters.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If traditional bundle-adjustment methods are used for 3D reconstruction, then computational efficiency is maintained, but accuracy in time-dependent scenes deteriorates

Engineering Contradiction:
Improveheight determination accuracyVSAvoidcomputational complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The scene is divided into static and dynamic components. Static points undergo traditional bundle-adjustment while dynamic points are handled separately with time-dependent height models. This segmentation allows accurate handling of time-dependent changes without applying complex computations to the entire scene, thus resolving the contradiction between measurement precision and computational complexity.

Inventive Principle:
Principle #1Segmentation

Solution Approach 2:

The invention introduces time as a dynamic parameter in the bundle-adjustment process. Heights of 3D points are modeled as time-dependent variables for dynamic regions, allowing the system to adapt to changes in the scene over time. This dynamic approach improves measurement precision for time-dependent scenes while maintaining computational efficiency by applying it only where necessary.

Inventive Principle:
Principle #15Dynamics

2Loss of information

If time-dependent height variations are allowed in bundle-adjustment, then identification of changing scene characteristics improves, but computational time increases

Engineering Contradiction:
Improveinformation about scene changesVSAvoidcomputational time
Core Design Contradiction:
Loss of informationVSLoss of time

Solution Approach 1:

The invention applies time-dependent height modeling locally only to dynamic regions of the scene rather than globally to all points. By identifying and separating dynamic from static regions, the system preserves information about scene changes in areas where it occurs while avoiding unnecessary computational overhead in static areas, thus resolving the contradiction between information retention and computational time.

Inventive Principle:
Principle #3Local quality

Solution Approach 2:

The system automatically identifies dynamic regions and applies appropriate processing only where needed. The bundle-adjustment process itself reveals which points exhibit time-dependent behavior, and the system adaptively applies time-dependent modeling only to those points, eliminating the need for manual specification and reducing overall computational time while preserving all relevant information.

Inventive Principle:
Principle #25Self-service

3Loss of information

If multiple images from different viewpoints and timings are processed, then completeness of scene information improves, but processing complexity increases

Engineering Contradiction:
Improvecompleteness of 3D informationVSAvoidprocessing system complexity
Core Design Contradiction:
Loss of informationVSDevice complexity

Solution Approach 1:

The processing system segments the bundle-adjustment task into two distinct phases: a first adjustment for static points and a second adjustment for dynamic points. This segmentation allows the system to process multiple images from different viewpoints and timings comprehensively while managing complexity by applying different processing strategies to different types of points, thus resolving the contradiction between information completeness and processing complexity.

Inventive Principle:
Principle #1Segmentation

Data Source

PatentEP4365839A1A computer implemented method for determining 3D coordinates describing a scene, a computer program, a computer program product, a device for determining 3D coordinates describing a scene and a display device for display of information relating to 3D coordinates describing a scene
Publication Date: 2024.05.08 MAXAR INT SWEDEN AB
  • EP4365839A1 patent drawingFigure 1
  • EP4365839A1 patent drawingFigure 2~6
  • EP4365839A1 patent drawingFigure 5a~5c

AI summary

The disclosure relates to a computer implemented method (100) for determining 3D coordinates describing a scene, obtaining (S1), (S3), S5) first, second and third images covering the scene captured from different viewpoints, said images being associated to a timing of capture, wherein the scene is formed in an overlapping geographical area of the first, second and third images. The method further comprises obtaining (S7) intrinsic and/or extrinsic parameter values for at least one imaging device used for capture of the images, and determining (S8) the 3D coordinates describing the scene, said determination comprising performing bundle-adjustments based on pair-wise measurements on the images to minimize a re-projection error of reconstructed 3D points in the overlapping geographical area, wherein time is used as a parameter in the determination of the 3D coordinates describing the scene, and wherein the re-projection error is minimized allowing at least a height to be locally time dependent in reconstruction of 3D points at least for a part of the overlapping geographical area. The disclosure further relates to a computer program, a computer program product, a device (200) for determining 3d coordinates describing a scene and a display device (300) for display of information relating to 3d coordinates describing a scene.