Aerial Vehicle Pose Localization Using Route Features Without GNSS
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Aerial vehicles rely on GNSS for navigation, which can be unreliable in environments with signal obstructions, and visual localization systems face challenges with environmental variability and feature recognition, affecting navigation accuracy and reliability.
Innovation Solution
The system uses absolute visual localization (AVL) to determine the pose of an aerial vehicle by analyzing route features with a neurally assisted surveying (NASR) compute and sensor module, integrating AVL-derived pose data with other sources and discarding unreliable data, and adjusting flight controls based on unified pose data.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If GNSS is used for navigation, then absolute position data is provided, but accuracy is reduced in environments with signal obstructions
Solution Approach 1:
The patent combines multiple navigation systems (GNSS, IMU, and visual localization system) into an integrated navigation solution. The processing system fuses data from these different sources to provide accurate position and orientation information, especially in environments where GNSS signals are obstructed. This merging allows the system to maintain high reliability and accuracy by compensating for the weaknesses of individual systems with the strengths of others.
Solution Approach 2:
The visual localization system acts as an intermediary between the aerial vehicle and the environment, providing alternative position and orientation data when GNSS is unavailable. By matching features from captured images with pre-stored map data, the system creates a reliable navigation reference that mediates the navigation problem in GNSS-denied environments.
2Reliability
If visual localization is used, then navigation without GNSS is enabled, but accuracy is affected by environmental variability and feature recognition challenges
Solution Approach 1:
The system performs preliminary actions by pre-capturing images of the route and storing them in a database before the actual navigation mission. These pre-stored images serve as reference data for feature matching during flight. By preparing the visual reference data in advance, the system ensures accurate feature recognition and matching during navigation, improving position accuracy even in variable environmental conditions.
Solution Approach 2:
The system implements feedback by continuously comparing features from real-time captured images with features from pre-stored map data. The processing system uses this comparison to calculate the aerial vehicle's position and orientation, and this information is fed back to adjust the navigation path. This closed-loop feedback mechanism improves measurement precision by continuously refining the position estimate based on visual feature matching accuracy.
Data Source
AI summary
The system and methods of the various embodiments may enable an aerial vehicle to determine its pose using absolute visual localization of route features and either a keypoint-based pipeline or a template based pipeline. This may result in the aerial vehicle being able to determine its pose when traditional methods of pose detection fail.