Road Boundary Estimation for Unmarked Drivable Surface Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional autonomous driving systems struggle to identify drivable surfaces outside marked lane boundaries, such as in areas without lane markings, construction zones, or when probe data is inaccurate.
Innovation Solution
The system receives satellite and probe data, extracts road features, determines lane groups, and generates a road boundary estimation network using a CNN to classify road sections and map trajectories for autonomous vehicles.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional autonomous driving systems rely on marked lane boundaries to identify drivable surfaces, then the system operation is simple and straightforward, but the system fails to identify drivable surfaces in areas without lane markings, construction zones, or when probe data is inaccurate
Solution Approach 1:
The system segments the road identification task into multiple components: satellite data processing, probe data processing, and fusion processing. By dividing the data sources and processing steps, the system can handle different road conditions (marked lanes, unmarked areas, construction zones) through appropriate data fusion strategies, improving both reliability and adaptability
Solution Approach 2:
The system merges satellite data and probe data to create a comprehensive view of drivable surfaces. This combination allows the system to cross-validate information and identify drivable areas even when one data source is insufficient or inaccurate, resolving the contradiction between reliability and adaptability
2Measurement precision
If the system uses only satellite data to extract road features, then the processing is simpler and faster, but the accuracy of drivable surface identification decreases in complex driving conditions
Solution Approach 1:
The system performs preliminary processing of satellite data and probe data separately before fusion. Satellite data is processed to extract road features in advance, and probe data is pre-processed to identify lane boundaries and traffic patterns. This preliminary action reduces the complexity of the fusion step while maintaining high precision in drivable surface identification
Solution Approach 2:
The system introduces an intermediary fusion processing step that combines satellite data and probe data. This intermediary layer reconciles the different data formats and resolutions, enabling high-precision identification without directly complexifying the individual data processing pipelines
3Measurement precision
If the system processes detailed probe data to determine lane groups, then the accuracy of trajectory mapping improves, but the processing time and computational resources increase
Solution Approach 1:
The system extracts only the essential features from probe data (lane boundaries, traffic patterns, road geometry) needed for trajectory mapping, rather than processing all raw probe data. This extraction approach maintains high mapping precision while significantly reducing processing time and computational resource requirements
Data Source
AI summary
Systems and methods are provided for mapping a trajectory for an autonomous vehicle. The system can receive satellite data and probe data of an autonomous vehicle traveling on a roadway and extract road features from the satellite data. Lane groups can be determined for sections of the roadway based on the probe data. The system can generate a road boundary estimation network to classify the sections of the roadway based on the road features and the lane groups and map a trajectory for the autonomous vehicle based on the road boundary estimation network.