Road Identification Using Trajectory and Lane Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for identifying the road traveled by a vehicle may inaccurately detect the road due to the precision of trajectory calculation and environmental factors, leading to incorrect information for autonomous driving control.
Innovation Solution
An apparatus and method that compare a vehicle's trajectory with a road map and sensor signals to identify the road, using confidence scores to ensure accurate detection, where the detected lane is included in or not included in the candidate road, and prioritizing the road with the highest matching score.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Extent of automation
If trajectory comparison with road map is used to identify the traveled road, then the identification process can be automated, but errors occur when the detected road does not match the actual traveled road due to trajectory accuracy or environmental factors
Solution Approach 1:
The patent combines two different road identification approaches: (1) trajectory-based road candidate detection, and (2) lane-based road detection using sensor signals. By merging these two methods and cross-validating their results, the system achieves more reliable road identification than either method alone could provide
Solution Approach 2:
The system uses feedback mechanisms where the detected lane information is used to verify and correct the trajectory-based road candidate. When the detected lane is not included in the candidate road, the system identifies a different road including the detected lane, thereby correcting potential errors through feedback validation
2Device complexity
If only trajectory information is used for road identification, then the system complexity is reduced, but the measurement precision of road identification deteriorates due to trajectory calculation errors
Solution Approach 1:
The patent segments the road identification task into two independent parts: trajectory-based road candidate identification and sensor-based lane detection. This segmentation allows each part to use the most appropriate method for its specific function, improving overall precision without requiring a single overly complex system
3Adaptability or versatility
If multiple road candidates are detected based on trajectory matching, then the adaptability of the system is improved, but the difficulty of detecting and measuring the correct road increases
Solution Approach 1:
The detected lane acts as an intermediary element that mediates between multiple road candidates and the final road identification. By using the lane detection results as an intermediary verification step, the system can efficiently select the correct road from multiple candidates without direct complex comparison between all candidates
Data Source
AI summary
An apparatus for identifying a road being traveled includes one or more processors configured to: compare a latest predetermined section of a trajectory of a vehicle with a first map representing roads to detect a section closest to a current position of the vehicle included in a road that matches the trajectory the best of the roads represented in the first map as a candidate for a road being traveled by the vehicle, compare a sensor signal representing environment around the vehicle with a second map representing individual lanes on roads to detect a lane being traveled by the vehicle, identify the candidate for the road as the road being traveled by the vehicle when the detected lane is included in the candidate, and identify a road including the detected lane as the road being traveled by the vehicle when the detected lane is not included in the candidate.