Inside-Out Tracking Map Refinement for Low-Visibility Environments
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional team location tracking and mapping systems are ineffective in high-stress, hazardous environments where external location services are unreliable and high-quality visible light images are difficult to capture.
Innovation Solution
An inside-out location tracking and mapping system that uses low-contrast and blurry images, such as those from thermal cameras, to perform dense feature tracking and refine maps through a method involving transform computation, appearance error evaluation, and greedy optimization.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional GPS and cellular location tracking systems are used, then location tracking works well in urban environments, but they become unreliable in remote locations and hazardous environments
Solution Approach 1:
The patent inverts the conventional outside-in location tracking approach by implementing inside-out tracking, where the tracking system is carried by the team members themselves rather than relying on external infrastructure. Each team member becomes a tracking node that can determine positions of others through visual features in the environment, enabling operation in remote areas without GPS or cellular coverage.
Solution Approach 2:
The patent introduces visual features in the environment as an intermediary medium for location tracking. Instead of direct satellite or cellular signals, the system uses visual features captured by cameras as mediators to estimate camera motion and determine relative positions between team members through image matching and geometric computation.
2Reliability
If camera-based inside-out tracking systems are used, then location tracking can work without external services, but they require high-quality visible light images which are difficult to capture in hazardous environments
Solution Approach 1:
The patent changes the operating parameters of the camera system by allowing operation with lower image quality thresholds. The system accepts images with lower contrast and higher motion blur than conventional systems would tolerate, and uses optimized algorithms to extract sufficient visual features from these degraded images for reliable tracking.
Solution Approach 2:
The patent makes the tracking system more tolerant of degraded input by creating a porous approach to feature extraction that can work with incomplete or low-quality visual data. The system can operate with sparse features and lower image quality, similar to how porous materials allow passage through restricted channels.
3Ease of operation
If sparse feature tracking is used, then location tracking can be implemented, but it requires high-quality images and geometric consistency that is difficult to maintain in low-visibility environments
Solution Approach 1:
The patent implements feedback mechanisms through map refinement processes that continuously evaluate and correct geometric consistency. The system computes appearance errors and refines the map based on accumulated tracking data, using feedback loops to maintain geometric consistency even when working with sparse features from low-quality images in challenging environments.
Data Source
AI summary
Techniques for map refinement (e.g., tracking data refinement) by an inside-out location tracking system may include computing a transform from a reference point to an epipolar line using an Essential Matrix derived from a reference frame camera motion in a live frame, computing several appearance errors between a feature associated with the reference point and other projected and optimized features, using a perpendicular projection hypothesis and an optimized point generated on the epipolar line, and evaluating the appearance errors using a greedy, ordered optimization. If the appearance error is less than a predetermined quality threshold, the optimized point is retained in tracking data. Otherwise the tracking data may be reset at the reference point location for reinitialization. An updated map and associated map data reflecting updated optimized points may be provided, for example, to a client device, as well as returned to a visual inertial odometry system.