Drivable Surface Polygon Generation for Accurate Road Graph Mapping
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The generation of road graphs is computationally complex, laborious, and inaccurate due to the manual annotation of vehicle sensor information, making it expensive.
Innovation Solution
A system and method that utilize key point detections and vehicle traces to generate drivable surface polygons, which are then used to create connection and separation graphs, followed by a three-dimensional mesh, to accurately represent road networks for electronic maps.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If manual annotation of vehicle sensor information is used to generate road graphs, then accuracy can be improved, but computational complexity and labor increase significantly
Solution Approach 1:
The patent segments the road graph generation process into distinct components: drivable surface polygon generation from sensor data, connection graph construction from polygons, and separation graph generation. This segmentation allows automated processing of each component, reducing manual annotation requirements while maintaining accuracy.
Solution Approach 2:
The system enables self-service by automatically generating drivable surface polygons from vehicle sensor information and traces without requiring manual annotation. The connection and separation graphs are then automatically constructed from these polygons, creating a fully automated pipeline that reduces both labor and computational complexity.
2Measurement precision
If manual annotation of vehicle sensor information is used to generate road graphs, then accuracy can be improved, but labor and cost increase significantly
Solution Approach 1:
The system performs self-service by automatically generating drivable surface polygons from vehicle sensor information and traces without manual intervention. The connection and separation graphs are automatically constructed from these polygons, eliminating labor-intensive manual annotation while maintaining road graph accuracy and significantly improving map generation efficiency.
Solution Approach 2:
The patent replaces the mechanical process of manual annotation with an automated computational system that processes vehicle sensor data and traces to generate drivable surface polygons and subsequent graph structures, thereby reducing labor requirements and increasing productivity.
3Productivity
If traditional road graph generation methods are used, then computational burden is high, but the systematic approach to representing road networks is limited
Solution Approach 1:
The patent segments the road network representation into three distinct data structures: drivable surface polygons that capture geometric boundaries, connection graphs that represent accessible road segments, and separation graphs that identify inaccessible areas. This segmentation improves computational efficiency while enhancing the system's ability to represent complex road network characteristics.
Solution Approach 2:
The patent introduces a new dimensional approach by generating drivable surface polygons as two-dimensional geometric representations from sensor data, which then serve as the foundation for constructing connection and separation graphs. This dimensional transformation enables more efficient computational processing while capturing comprehensive road network information.
Data Source
AI summary
Systems and methods for generating drivable surface polygons for electronic map generation are described herein. In one example, a system includes a processor and a memory in communication with the processor having instructions that, when executed by the processor, cause the processor to generate drivable surface polygons that represent a drivable surface of a road using key point detections and a vehicle trace and generate an electronic map based on the drivable surface polygons.