Concurrent Track Association Using Graph Crossings
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional machine vision systems lack inherent support for continuous surface representation of objects, and it is difficult to efficiently determine if scanning information from different sensors represent the same points or objects in the scanned subject matter.
Innovation Solution
The association of concurrent tracks using graph crossings is achieved by employing two or more signal beams to scan an object, generating trajectories based on detected signals, determining crossing points, and comparing graphs to identify common trajectories across the object, thereby determining object features.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multiple scanning sensors are used to concurrently scan the same scene, then the quantity and coverage of scanning information is improved, but the difficulty of determining if scanning information from different sensors represent the same points or objects increases
Solution Approach 1:
The patent introduces a temporal signature as an intermediary marker that is encoded into the scanning information by each sensor. This temporal signature acts as a mediator that allows the system to automatically associate scanning information from different sensors by comparing their temporal signatures, thereby resolving the difficulty of determining whether scanning information represents the same points or objects without requiring complex cross-sensor analysis
Solution Approach 2:
The system employs feedback by continuously monitoring and comparing temporal signatures from multiple sensors. When scanning information with matching temporal signatures is detected, the system provides feedback to confirm that the information represents the same physical point or object, enabling efficient association of concurrent tracks from multiple sensors
2Device complexity
If conventional 2D or 3D representations are used, then the simplicity of data structure is maintained, but the ability to provide continuous surface representation of objects is lost
Solution Approach 1:
The patent adds a temporal dimension to the traditional 2D or 3D representations by incorporating temporal signatures that encode continuous surface information. This transforms static point cloud or image data into temporally-aware representations that can convey continuous surface properties while maintaining compatibility with conventional data structures through the addition of temporal metadata
Data Source
AI summary
Embodiments are directed to the association of concurrent tracks using graph crossings. Signal beams may be employed to scan paths across an object such that sensors separately detect signals from the signal beams reflected by the object. Crossing points may be determined based on a plurality of trajectories that intersect each other during the scan of the object. Graphs may be generated based on the portion of trajectories and the crossing points such that each edge in the graphs corresponds to a crossing point and such that each node in the one or more graphs corresponds to a trajectory. The graphs may be compared to determine one or more matched graphs that may share a common topology. Common trajectories may be determined based on the matched graphs such that each common trajectory may be associated with a same path across the object and a separate sensor.