Aerial Vehicle Pose Localization Using Visual Route Features
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Traditional navigation systems for aerial vehicles, such as drones, rely on GNSS and IMU, which are susceptible to signal interference and drift, leading to inaccurate pose estimation in complex environments.
Innovation Solution
Utilizing absolute visual localization (AVL) to determine pose by analyzing route features with onboard cameras, integrating with IMU and GNSS data for enhanced navigation, and employing neural networks for image processing to improve accuracy.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If GNSS and IMU are used for navigation, then positioning is provided, but accuracy deteriorates in complex environments due to signal obstructions and drift
Solution Approach 1:
The patent introduces visual landmarks as an intermediary reference system between the aerial vehicle and the environment. By detecting and matching visual features in the environment against a pre-built map, the system establishes a reliable pose estimation mechanism that does not depend on GNSS signals or IMU integration, thereby resolving the accuracy problem in complex environments.
Solution Approach 2:
The patent replaces the mechanical/sensor-based navigation system (GNSS+IMU) with a vision-based localization system. Instead of relying on inertial measurements and satellite signals that suffer from drift and signal loss, the system uses image processing and feature matching to determine pose, substituting the navigation approach to eliminate the fundamental accuracy limitations.
2Adaptability or versatility
If multiple sensor modalities are integrated, then navigation coverage is improved, but system complexity increases
Solution Approach 1:
The patent extracts and focuses solely on vision-based localization for pose estimation, separating this function from the traditional GNSS/IMU navigation stack. By taking out the visual localization component as a distinct system that operates independently, the patent achieves navigation coverage in GNSS-denied areas without requiring complex integration of multiple sensor modalities, thereby reducing overall system complexity.
Data Source
Figure 1
Figure 2
Figure 3A
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.