Propagating Traffic Sign Data Across Road Links
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems face challenges in ensuring comprehensive traffic sign data coverage in digital maps due to gaps in data, particularly on minor roads where fewer vehicles travel and report observations, leading to incomplete and inaccurate information for autonomous driving applications.
Innovation Solution
A method and apparatus for propagating learned traffic sign data from upstream links to downstream links based on road attributes and connectivity, ensuring that all links in a road network are labeled with traffic sign information, even if no signs are directly detected on them.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If learned traffic sign data is only assigned to road links where signs are directly detected, then data accuracy is maintained, but data coverage is incomplete particularly on minor roads
Solution Approach 1:
The patent applies preliminary action by propagating traffic sign data from upstream road links to downstream road links before actual detection is needed. This allows minor roads without direct detections to receive traffic sign data in advance through propagation from connected major roads, thereby improving coverage while maintaining accuracy through attribute-based filtering
Solution Approach 2:
The patent uses road link attributes (functional class, speed limit, geometry) as intermediaries to bridge the gap between detected and undetected road links. These attributes act as mediators to determine whether traffic sign data should be propagated from upstream to downstream links, enabling accurate coverage extension without direct detection
2Loss of information
If more vehicles are deployed to report traffic sign observations, then data coverage improves, but system cost and complexity increase
Solution Approach 1:
The patent implements self-service by enabling the system to automatically fill data gaps through propagation algorithms rather than relying on additional vehicles. The system uses existing detected data and road attributes to autonomously populate traffic sign information for undetected links, eliminating the need for expanded vehicle fleets while maintaining coverage
Solution Approach 2:
The patent applies copying by replicating traffic sign data from upstream road links to downstream road links based on attribute similarity. Instead of deploying more vehicles to detect signs on every link, the system creates accurate copies of sign data through propagation, using road attributes as validation criteria to ensure the copied data is appropriate for the target link
Data Source
AI summary
An approach is provided for propagating learned traffic sign data. The approach involves, for example, determining a road link to which learned traffic sign data has been assigned. The approach also involves identifying one or more downstream links connected to the road link to which no learned traffic sign data has been assigned. The approach further involves propagating the learned traffic sign data of the road link to the identified one or more downstream links.