Event-Camera Airborne Navigation for GPS-Denied Relative Motion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing navigation methods for airborne platforms, such as satellite-based systems and image-based location determination, face challenges in accuracy and computational intensity, especially in environments where satellite signals are obstructed or jammed, and require significant processing power and memory.
Innovation Solution
A navigation method using an event camera on an airborne platform that captures intensity gradient changes to determine relative motion by comparing signatures of ground-referenced intensity gradients between data capture cycles, allowing for efficient relative navigation without or with reduced reliance on inertial management units.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If image-based location determination is used to determine platform location in a large possible search space, then location accuracy can be achieved, but computational intensity and processing time increase significantly
Solution Approach 1:
The patent segments the image-based navigation problem into two parts: (1) using an event camera to capture only intensity gradient change events rather than full images, and (2) using these sparse events to compute intensity gradient directions and distributions. This segmentation reduces the data volume and computational complexity while maintaining the ability to determine platform location and orientation.
Solution Approach 2:
The patent extracts only the essential information needed for navigation from the visual data - specifically, intensity gradient change events and their directions. By taking out only these critical features rather than processing complete images, the system achieves location determination with significantly reduced computational intensity and memory requirements.
2Measurement precision
If conventional image-based location determination is used, then location can be determined, but significant memory is required to store database of features
Solution Approach 1:
The patent extracts only intensity gradient event data from the visual scene, storing minimal information about event locations, directions, and intensities. This extracted event data replaces the need for storing large databases of reference images or features, dramatically reducing memory requirements while maintaining location determination capability.
Solution Approach 2:
The patent creates a simplified representation (copy) of the visual environment through intensity gradient events rather than storing actual image data or feature databases. This event-based copy contains sufficient information for navigation but occupies minimal memory space.
3Measurement precision
If satellite-based navigation systems are used, then absolute location can be determined, but signals may be obstructed or jammed
Solution Approach 1:
The patent uses intensity gradient events as an intermediary between the camera and navigation computation. These events serve as a robust intermediate representation that can be reliably captured and processed even when satellite signals are unavailable, enabling continuous navigation functionality.
Solution Approach 2:
The system performs self-service navigation by using its own camera to capture intensity gradient events and compute platform motion and location independently of external satellite signals. This self-contained approach ensures reliable operation in environments where satellite signals are obstructed or jammed.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables accurate and computationally efficient relative navigation by identifying overlap areas between data capture cycles, reducing drift, and enabling lower quality inertial units, suitable for real-time navigation.
Implementation Method 1
recording events triggered during motion of the sightline, each event being a change in intensity recorded at a pixel of the camera
Data Source
Figure 1~2
Figure 3~4
Figure 5a~5b
AI summary
There is disclosed a navigation method for an airborne platform. The airborne platform has a ground-facing camera having a field of view, and the method comprises repeating a data capture cycle. The data capture cycle comprises the steps of: moving the sightline of the camera through a range of directions; recording events triggered during motion of the sightline, each event being a change in intensity recorded at a pixel of the camera, and associating each event with a ground-referenced position; and, for each ground-referenced position, determining, from the recorded events, a resultant intensity gradient direction. Relative motion of the platform is determined by comparing a distribution of resultant gradient intensities obtained in a current data capture cycle with a distribution of resultant gradient intensities obtained in an earlier data capture cycle.