Stereo Camera Positioning via Reference Point Correlation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing visual survey systems face measurement errors due to physical limitations, particularly in determining the accurate position and orientation of objects over distance, which affects the accuracy of lane departure warning systems, vehicle automation, and infrastructure mapping.
Innovation Solution
A positional computation system and method that uses a feedback technique with imagery data from a stereo camera pair to correct the relative position of nearby survey capture points, creating a 'rigid' mesh to minimize errors by correlating reference points and averaging over greater sample sets.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Area of stationary object
If visual survey systems are used to determine object position and orientation over distance, then survey coverage area is increased, but measurement precision deteriorates due to physical limitations of the survey systems
Solution Approach 1:
The patent divides the survey area into multiple overlapping image capture zones. By segmenting the overall survey into discrete image sets with known overlap regions, the system can process smaller localized areas with higher precision while maintaining extensive overall coverage. The segmentation allows reference points to be correlated within specific overlap regions rather than across the entire survey area.
Solution Approach 2:
The patent introduces reference points as intermediary elements that appear in overlapping image sets. These reference points serve as mediators to correlate and integrate data from different image captures. By using reference points as intermediaries, the system can determine absolute positions and orientations more accurately while maintaining broad survey coverage through the network of overlapping images.
2Area of stationary object
If multiple measurements are taken to determine object position, then survey coverage is improved, but measurement errors are multiplied through several measurements
Solution Approach 1:
The patent implements a feedback mechanism where reference points identified in overlapping image sets are used to correct and refine position determinations. The system correlates reference points across multiple images, uses these correlations to determine absolute positions and orientations, and feeds this information back to reduce errors in subsequent measurements. This feedback loop prevents error multiplication by continuously refining measurements based on correlated reference data.
Solution Approach 2:
The patent performs preliminary correlation of reference points in overlapping image sets before final position determination. By pre-identifying and correlating reference points across multiple images, the system establishes a framework of known reference positions that can be used to guide and constrain subsequent object position calculations, thereby preventing error accumulation rather than allowing it to propagate through multiple measurements.
3Measurement precision
If reference points are correlated between overlapping image sets, then object location accuracy is improved, but computational complexity increases
Solution Approach 1:
The patent applies local quality by focusing computational resources on correlating reference points within specific overlapping regions rather than processing the entire survey area uniformly. Each overlap region is processed with high precision to establish local reference point correlations, which then serve as the basis for broader position determinations. This localized approach improves object location accuracy while managing computational complexity by avoiding redundant global processing.
Data Source
AI summary
The present invention involves a surveying system and method which determines the position of an object point using two images. First, at least two reference points appearing on the two images are correlated. Then the position of the object point is determined based on the two images and the two reference points.