Lane Change Detection via Graph Distance Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing lane change detection systems in vehicles rely on complex and costly solutions that involve direct observation of the road ahead, which may not be feasible for cost-sensitive vehicle platforms.
Innovation Solution
A lane change detection system that uses a graph representation to determine left-handed and right-handed distances based on vehicle position and lane boundary indicators, evaluating changes in magnitude to detect lane changes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If complex sensor systems (cameras, LiDAR) are used for lane change detection, then detection accuracy is improved, but system cost increases
Solution Approach 1:
The patent creates a graph representation (copy) of the road geometry and vehicle position based on existing sensor data, rather than using expensive dedicated sensors. This virtual model allows lane change detection through computational analysis of distance relationships in the graph, achieving accurate detection without additional costly hardware
Solution Approach 2:
The patent replaces physical sensor systems (cameras, LiDAR) with a computational graph-based approach. Instead of using mechanical/optical sensors to directly detect lane changes, the system uses mathematical graph representations and distance calculations to infer lane change events, substituting hardware complexity with software-based detection
2Reliability
If direct observation of road ahead is used for lane change detection, then detection reliability is improved, but system complexity increases
Solution Approach 1:
The patent transforms the lane change detection problem from direct spatial observation into a graph-theoretic dimension. By representing road geometry, vehicle position, and lane boundaries as nodes and edges in a graph structure, the system detects lane changes through topological and metric relationships in the graph space rather than direct visual observation, simplifying the detection logic while maintaining reliability
Data Source
AI summary
System, methods, and other embodiments described herein relate to implementing lane change detection strategies. In one embodiment, a method includes determining for each segment of a frame graph a left-handed distance and a right-handed distance based on a vehicle position indicator, a set of left lane boundary indicators, and a set of right lane boundary indicators within each segment; and determining if a lane change has occurred by evaluating changes in magnitude within a set of left-handed distances and a set of right-handed distances.