Sign Classification Using Blocking Link Verification
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current navigation applications face inaccuracies in mapping due to onboard sensors mistakenly capturing signs from parallel pathways, leading to erroneous map data and potential navigation errors, including accidents.
Innovation Solution
A system and method for generating classification data of signs by determining the presence of blocking links between map-matched and parallel links based on heading and distance criteria, ensuring accurate association of signs with the correct link.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If onboard sensors capture signs with wide field-of-view, then more signs can be detected, but signs from parallel pathways are mistakenly captured leading to mapping errors
Solution Approach 1:
The patent introduces an intermediary blocking link analysis between the sign detection and final classification. By examining whether blocking links exist between parallel pathways, the system mediates the classification decision to prevent misattribution of signs from parallel paths to the current pathway, thus resolving the contradiction between wide detection coverage and accuracy.
2Ease of operation
If signs are classified based on proximity to current pathway, then classification is simple, but signs from parallel pathways are incorrectly associated
Solution Approach 1:
The blocking link serves as an intermediary verification mechanism. Instead of directly associating signs with the nearest pathway, the system checks for the presence of blocking links that would prevent line-of-sight between the sign and parallel pathways. This adds a reliable verification step while maintaining operational feasibility through automated geometric analysis.
Solution Approach 2:
The patent replaces complex manual verification of sign associations with automated geometric analysis of blocking links. By using computational geometry to determine whether links block visibility between signs and parallel pathways, the system substitutes sophisticated automated reasoning for simple proximity-based classification, improving reliability without significantly increasing operational complexity.
3Productivity
If map data is built from learned signs, then navigation assistance is provided, but inaccurate map data leads to navigation errors and accidents
Solution Approach 1:
The patent performs preliminary classification verification before incorporating signs into map data. By checking for blocking links during the sign classification phase, the system prevents erroneous data from entering the map database in the first place. This preliminary action ensures that only accurately classified signs are used for navigation assistance, maintaining both productivity and reliability.
Data Source
AI summary
A system, a method, and a computer program product for generating classification data of a sign. The method comprises obtaining map data of the sign, obtaining a map matched link and at least one parallel link associated with the sign based on the map data, determining presence data of at least one blocking link between the map-matched link and the at least one parallel link, wherein the at least one blocking link satisfies one or more of a heading criterion and a distance criterion. The method further comprises generating the classification data of the sign based on the presence data of the at least one blocking link.