Optical Flow Uncertainty Mapping for Vehicle Pose Estimation

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Conventional techniques fail to utilize uncertainty maps to determine the reliability of optical flow output, lacking efficiency and accuracy in estimating pose, velocity, and attitude of unmanned aerial vehicles and urban air mobility vehicles.

Innovation Solution

The solution employs a method utilizing an uncertainty map to select a best set of optical flow estimates for one or more vehicles, which includes receiving images captured by one or more cameras of a vehicle, analyzing the images by inputting the one or more processors, one or more processors, and determining a feature map based on the optical flow map and uncertainty map indicating a reliability of the optical flow map.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Measurement precision

If conventional techniques are used to estimate pose, velocity, and attitude, then the estimation process is simple, but the accuracy and reliability are insufficient due to lack of uncertainty assessment

Engineering Contradiction:
Improveestimation accuracyVSAvoidsystem complexity
Core Design Contradiction:
Measurement precisionVSDevice complexity

Solution Approach 1:

The patent introduces an uncertainty map as an intermediary component that assesses the reliability of optical flow estimates. This uncertainty map acts as a mediator between the optical flow estimation process and the final pose/velocity/attitude calculation, allowing the system to weight or discard estimates based on their reliability without fundamentally changing the core estimation algorithms.

Inventive Principle:
Principle #24Intermediary (Mediator)

Solution Approach 2:

The patent replaces traditional mechanical or algorithmic filtering methods with a data-driven uncertainty assessment approach. Instead of using fixed thresholds or complex mechanical filtering systems, the invention uses learned uncertainty patterns from the optical flow process itself to dynamically adjust estimation reliability, substituting mechanical complexity with intelligent algorithms.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

2Measurement precision

If multiple optical flow estimates are processed to improve accuracy, then the estimation precision increases, but the processing time and computational load increase

Engineering Contradiction:
Improveestimation precisionVSAvoidprocessing time
Core Design Contradiction:
Measurement precisionVSLoss of time

Solution Approach 1:

The patent extracts and utilizes the uncertainty information that is already generated during the optical flow estimation process. Instead of processing multiple complete optical flow estimates through full analysis pipelines, the invention extracts uncertainty metrics from each estimate and uses these to selectively process only the most reliable estimates, thereby reducing overall processing time while maintaining or improving precision.

Inventive Principle:
Principle #2Taking out (Extraction)

Solution Approach 2:

The patent applies partial processing to optical flow estimates based on their uncertainty levels. Rather than fully processing all estimates equally, the system performs partial processing on high-uncertainty estimates (using them only when necessary) and full processing on low-uncertainty estimates, optimizing the balance between precision and processing time by avoiding excessive computation on already reliable data.

Inventive Principle:
Principle #16Partial or excessive action

Data Source

PatentEP4421451B1An effective method to estimate pose, velocity and attitude with uncertainty
Publication Date: 2026.04.15 HONEYWELL INTERNATIONAL INC
  • EP4421451B1 patent drawingFigure 1
  • EP4421451B1 patent drawingFigure 2
  • EP4421451B1 patent drawingFigure 3

AI summary

Disclosed are methods and systems for utilizing an uncertainty map to select a best set of optical flow estimates for one or more vehicles. For instance, a method may include receiving one or more images captured by one or more cameras of at least one vehicle, analyzing the one or more images by inputting the one or more images into a trained machine-learning model, wherein the trained machine-learning model is configured to process the one or more images to determine an optical flow map and an uncertainty map, determining a feature map based on the optical flow map, determining an improved feature map by applying the uncertainty map to the feature map, transforming the improved feature map by applying at least one transformation or optimization to the improved feature map, and based on the transformed improved feature map, estimating a pose value, a velocity value, or an attitude value.