SD and HD Map Alignment via Graph Matching
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current methods fail to effectively cross-reference and align Standard Definition (SD) maps with High Definition (HD) maps from different sources, limiting the ability to connect and update road data accurately across different navigation systems.
Innovation Solution
An algorithm that aligns SD and HD maps by generating a graph from HD map data and converting SD waypoints to HD nodes, enabling the retrieval of HD segment IDs and lane-level details for autonomous driving systems, using incremental spatial joins and data from external sources like Traffic Management Channel codes.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a static lookup table with one-to-one relationship between SD road segments and HD road segments is maintained, then cross-referencing between SD and HD maps is enabled, but the system cannot adapt to changes from external sources and lacks versatility
Solution Approach 1:
The patent transforms the static lookup table into a dynamic alignment system that continuously adapts to changes in both SD and HD map data. The algorithm processes changes from external sources (like TMC codes) and recalculates alignments, ensuring the system remains reliable while adapting to new data versions and sources.
Solution Approach 2:
The alignment algorithm is designed to work with multiple data sources and formats, including TMC codes, SD map changes, and HD map updates. This universal approach enables cross-referencing between different map providers and formats, enhancing versatility while maintaining data accuracy through standardized processing.
2Measurement precision
If SD waypoints are converted to HD nodes using graph generation, then precise road and lane-level information is retrieved, but the computational complexity and processing time increase
Solution Approach 1:
The patent segments the HD map into a graph structure where nodes represent key locations and edges represent road segments. This segmentation allows the system to process and query specific portions of the map independently, reducing overall computational complexity while maintaining precise alignment through localized graph operations.
Solution Approach 2:
The graph structure serves as an intermediary between SD waypoints and HD map data. By converting SD waypoints to HD nodes through this intermediate graph representation, the system achieves precise alignment without directly comparing all SD and HD data elements, thereby reducing computational complexity.
3Reliability
If the alignment algorithm processes changes from external sources, then data accuracy is maintained, but the processing time and computational resources increase
Solution Approach 1:
The patent pre-processes and stores map data in structured formats (graph structures, lookup tables) before changes occur. When external changes arrive (such as TMC updates), the system can quickly apply these changes to the pre-processed data structures, maintaining data accuracy while minimizing processing time compared to processing raw data from scratch.
4Loss of information
If HD map data is loaded into memory for autonomous driving systems, then detailed navigation information is available, but the memory footprint increases
Solution Approach 1:
The patent extracts only the necessary HD map data segments that are relevant to the current route and autonomous driving needs. By using the alignment algorithm to identify and extract specific road segments, lanes, and features from the full HD map, the system maintains complete navigation information while significantly reducing memory footprint by loading only essential data.
Data Source
AI summary
Aspects of the disclosed invention relate to alignment of standard definition (SD) maps and high definition (HD maps) which may come from different sources. Responsive to input of a destination, a route to that destination may be defined, and SD map waypoints generated from that defined route. A graph may be generated from the HD map. The waypoints may be matched with nodes and edges in the graph. One or more edges may constitute a segment in the HD map. A plurality of segments are identified to match the route.