Vehicle Self-Location Estimation Using Traffic Sign Height
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle location determination technologies cannot accurately differentiate between roads of the same type at different heights, especially when their two-dimensional locations are close, leading to uncertainty about the vehicle's position.
Innovation Solution
A self-location estimation method using a driving assistance device that incorporates a camera, ranging device, and map database to estimate the vehicle's height by detecting traffic signs and correcting dead reckoning errors with height information from the map database, allowing for accurate determination of the road on which the vehicle is present without requiring three-dimensional map matching.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Device complexity
If two-dimensional map matching is used to determine vehicle location, then computational complexity is reduced, but the ability to differentiate between roads at different heights is lost
Solution Approach 1:
The patent introduces height information as a third dimension to resolve ambiguity in two-dimensional map matching. By detecting the height of traffic signs relative to the vehicle and comparing it with map database information, the system can differentiate between roads at different heights (e.g., overpasses and ground-level roads) that appear identical in 2D space, thereby maintaining location determination accuracy without requiring full three-dimensional map matching.
2Measurement precision
If three-dimensional map matching is used to determine vehicle location on roads at different heights, then location determination accuracy is improved, but computational complexity and error accumulation increase
Solution Approach 1:
The patent extracts only the necessary height information from three-dimensional map matching by focusing specifically on traffic sign height detection. Instead of performing full 3D map matching, the system detects the height of traffic signs using cameras or ranging devices, compares this with pre-stored height information in the map database, and uses this extracted height parameter to determine the vehicle's vertical position. This selective extraction maintains accuracy while reducing computational burden.
Solution Approach 2:
The patent uses traffic signs as intermediary objects to bridge the gap between 2D and 3D location determination. Traffic signs serve as reference objects with known heights stored in the map database. By detecting these intermediaries and their relative heights to the vehicle, the system can infer the vehicle's height above ground level without requiring complex 3D spatial reasoning, thus simplifying the overall computation while maintaining precision.
3Device complexity
If dead reckoning is used for self-location estimation, then computational cost is reduced, but error accumulation occurs over time
Solution Approach 1:
The patent implements feedback by using detected traffic sign heights to correct accumulated errors in dead reckoning estimation. The system continuously compares the height-derived location information with the dead reckoning results and adjusts the estimated position accordingly. This feedback mechanism prevents error accumulation while maintaining the low computational cost advantage of dead reckoning, as the corrections are applied periodically rather than requiring continuous complex calculations.
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A self-location estimation method including: detecting a traffic sign in a vicinity of a vehicle (S12); determining a current location of the vehicle (S11); acquiring height information of the detected traffic sign from map data in which two-dimensional coordinate information and height information of a traffic sign existing in a vicinity of a road are recorded, based on a relative location of the detected traffic sign with respect to the vehicle and the current location; and estimating a height at which the vehicle is present according to the height information acquired from the map data (S13-S15).