Intersection Center Localization Using Traffic Light Geometry
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Advanced driving assist and autonomous driving technologies face challenges in detecting the configuration of intersections and lane-junction relationships due to high costs associated with high-definition maps and difficulties in recognizing lane markers and road edges using autonomous sensors.
Innovation Solution
An apparatus and method that utilize a surrounding monitor to obtain data, identify the center of a road, detect traffic light devices, and calculate an intersection point between the road center and a line connecting traffic light devices to localize the intersection center, thereby determining the type of lane the vehicle is on.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If high-definition maps are used to detect intersection configuration and lane-junction relationships, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent uses traffic light devices as intermediary objects to detect intersection centers. Instead of directly using high-definition maps, the system detects traffic light positions and uses them as mediators to calculate intersection centers through geometric relationships, thereby reducing dependency on complex map data while maintaining detection accuracy
Solution Approach 2:
The patent replaces the mechanical/system-dependent approach of using high-definition maps with an optical/geometric approach. By using camera-based traffic light detection and geometric calculations (line intersections) to determine intersection centers, the system substitutes map-based mechanical positioning with visual-field-based geometric reasoning
2Ease of operation
If autonomous sensors are used to recognize lane markers and road edges, then ease of operation is improved, but measurement precision deteriorates
Solution Approach 1:
The patent introduces traffic light devices as intermediary reference points to improve lane recognition accuracy. By detecting traffic light positions and using them to define intersection centers and lane boundaries, the system creates more reliable reference points than relying solely on autonomous sensor detection of lane markers and road edges
3Device complexity
If traffic light devices are used to localize intersection center, then device complexity is reduced, but measurement precision may worsen
Solution Approach 1:
The patent transitions from two-dimensional map-based intersection detection to three-dimensional spatial reasoning by using traffic light device positions in the visual field. By calculating the intersection point of lines connecting traffic light positions with the vehicle position, the system creates a new dimensional approach to localization that maintains accuracy while reducing complexity
Solution Approach 2:
The patent creates a simplified geometric model (copy) of the intersection environment using only essential elements: vehicle position and traffic light positions. This copied geometric representation allows the system to calculate intersection centers without needing complete and complex environmental data, thereby reducing device complexity while maintaining sufficient measurement precision
Data Source
AI summary
In an apparatus for localizing a center of an intersection, a data obtainer obtains, from a surrounding monitor that monitors a surrounding situation around a vehicle traveling on a road, data indicative of the surrounding situation. A road center identifier identifies a location of a center of the road based on the data indicative of the surrounding situation around the vehicle. A traffic light detector detects, based on the data indicative of the surrounding situation around the vehicle, at least two traffic light devices located in an intersection to which the road joins. A center localization unit detects an intersection point between the location of the center of the road identified by the road center identifier and a line connecting between the at least two traffic light devices detected by the traffic light detector, thus localizing the detected intersection point as a center of the intersection.