Multi-View Track Association for 3D Point Triangulation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional machine vision systems struggle to provide 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
A sensing system employing two or more signal beams to scan a plurality of paths across an object, using two or more sensors to detect reflected signals, generating essential matrices based on sensor positions, comparing events, and associating them based on error scores to triangulate three-dimensional positions.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multiple scanning sensors are employed to concurrently scan the same scenes, 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 an association engine as an intermediary component that receives scanning information from multiple sensors and processes it to determine correspondences. The association engine uses essential matrices and error score calculations as intermediate computational steps to bridge the gap between raw sensor data and meaningful associations, making the complex multi-sensor integration manageable and systematic
Solution Approach 2:
The system implements a feedback mechanism where the association engine calculates error scores for potential point associations and uses these error scores to iteratively refine and confirm correspondences. This feedback loop allows the system to validate associations and adjust its determination of whether points from different sensors represent the same physical locations
2Ease of operation
If conventional machine vision algorithms are used to process 2D or 3D data, then basic object detection is achieved, but inherent support for continuous surface representation is lacking
Solution Approach 1:
The patent transitions from processing discrete 2D image data or unstructured 3D point clouds to representing objects as continuous surfaces in three-dimensional space. By introducing a coordinate system and surface representation model, the system adds the dimension of continuous spatial representation, enabling more versatile robotic manipulation tasks that require understanding of object geometry and surface properties
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 efficient association of concurrent tracks across multiple views, providing accurate three-dimensional positioning of objects by triangulating point locations based on sensor events.
Implementation Method 1
two or more sensors to detect reflected signals
Implementation Method 2
triangulate three-dimensional positions
Data Source
AI summary
Embodiments are directed to the association of concurrent tracks across multiple views for sensing objects. A sensing system that employs signal beams to scan paths across an object may be provided such that two or more sensors separately detect signals reflected by the object. Events may be determined based on the detected signals such that the events include pairs of events that correspond to pairs of sensors. Essential matrices may be generated based on the positions of each pair of sensors. The pairs of events associated with the pairs of sensors may be compared based on the essential matrices of the sensors. Scores for the pairs of events may be provided based on the comparison. If a score for a pair of events may be less than a threshold value, each event in the pair of events may be associated with a same location on the object.