Intersection Road Edge Geometry for Occlusion-Robust Vehicle Navigation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Detecting road edges at intersections is challenging due to issues like faded markings, low-quality images, and occlusions, and crowd-sourced data methods are prone to noise and data sparsity, making it difficult to construct accurate road edges for autonomous vehicle navigation.
Innovation Solution
A method and system that determine the first and second road edges entering an intersection, construct an intersection edge by connecting points on these edges, and transmit this information to vehicles for navigation, using a remote processor and map server to calculate a nominal turn radius, tangent distance, and turn center, allowing for interpolation of an intersection radial line.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If aerial imagery is used for road edge inference, then road edges can be detected, but detection accuracy deteriorates due to faded markings, low quality images, and occlusions
Solution Approach 1:
The patent uses road centerlines as an intermediary element to indirectly determine road edges at intersections. Instead of directly detecting obscured edges from aerial imagery, the system traces road centerlines from approach roads through the intersection and uses geometric relationships to infer edge positions, effectively using the centerline as a mediator that bypasses the occlusion problem
Solution Approach 2:
The system performs preliminary detection of road centerlines and edges on approach roads before reaching the intersection. By pre-establishing these geometric references and extending them through the intersection using geometric modeling, the system prepares the necessary spatial information in advance to accurately construct intersection edges without relying on degraded aerial imagery at the intersection itself
2Quantity of substance
If crowd-sourced vehicle data is used to detect road edges, then more data can be collected, but data quality deteriorates due to noise and data sparsity
Solution Approach 1:
The patent extracts and utilizes only the essential geometric information (road centerlines and edge points) from crowd-sourced vehicle data, discarding noisy or redundant information. By focusing on extracting clean geometric features rather than processing all raw sensor data, the system obtains sufficient information for accurate edge construction without being affected by data noise and sparsity
Solution Approach 2:
The system creates a simplified geometric representation (copy) of the road network using extracted centerline and edge point data. This geometric model serves as a cleaned, noise-free version that captures the essential road structure without the imperfections present in raw crowd-sourced sensor data
3Reliability
If map-matching efforts are performed for crowd-sourced data, then road edges can be detected, but system complexity increases and costs increase
Solution Approach 1:
The system performs self-service by using its own geometric modeling capabilities to directly construct road edges at intersections from extracted centerline data, without requiring external map-matching services. The geometric relationships and intersection models enable the system to independently determine edge positions, eliminating the need for complex map-matching infrastructure and reducing system complexity
Data Source
AI summary
A system, map server and method of navigating a vehicle through an intersection. The map server includes a remote processor and a communication device. The remote processor determines a first road edge for a first road entering an intersection and a second road edge for a second road entering the intersection. The remote processor constructs an intersection edge that connects a first point on the first road edge to a second point on the second road edge. The communication device communicates the intersection edge to the vehicle. A vehicle processor at the vehicle navigates the vehicle through the intersection using the intersection edge.