Road Network Generation from GPS Trajectories
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing road map inference technologies using GPS trajectory data are inefficient and not scalable, requiring significant computation for fully functional maps that include unnecessary traffic information for navigation purposes, making them redundant for applications like trajectory pattern mining and vehicle fleet management.
Innovation Solution
A method and apparatus for generating a road network by aggregating grid cells on a trajectory map to form level-1 regions and merging valid neighbors to create links, reducing data size and computation overhead, allowing for more efficient and scalable map inference.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If fully functional road maps are generated using GPS trajectory data, then navigation functionality is achieved, but computation overhead increases significantly and scalability is poor
Solution Approach 1:
The patent extracts only the essential geometric information from GPS trajectories to create a simplified road network representation. It removes unnecessary navigation-related data (such as turn directions, traffic conditions, and detailed route instructions) while retaining the core structural information needed for spatial analysis, thereby reducing computation requirements
Solution Approach 2:
The patent segments the complex process of road map inference into distinct modules: trajectory data processing, grid cell aggregation, and road network generation. This segmentation allows each component to be optimized independently and enables progressive processing of large datasets, improving overall computational efficiency
2Adaptability or versatility
If fully functional road maps are generated, then complete navigation information is obtained, but data size increases making transmission and storage inefficient
Solution Approach 1:
The patent extracts and retains only the geometric structure information (road segments and intersections) while discarding redundant navigation-specific data. This results in a minimal representation that captures the essential spatial topology without the bulk of unnecessary navigation metadata, significantly reducing data volume
Solution Approach 2:
The patent merges multiple GPS trajectories into a unified road network representation by aggregating grid cells and merging overlapping segments. This consolidation process combines information from multiple data sources into a single compact structure, reducing total data size while preserving spatial accuracy
3Measurement precision
If individual GPS trajectory geometrical features are computed for fully functional maps, then accurate navigation routing is achieved, but computation time increases significantly
Solution Approach 1:
The patent performs preliminary processing by pre-dividing the study area into grid cells and pre-aggregating trajectory data into these grids before generating the final road network. This preliminary action simplifies subsequent processing steps and reduces the computational time required for accurate routing calculations
Solution Approach 2:
The patent creates a simplified geometric representation (road network graph) that copies only the essential topological relationships from the complex GPS trajectory data. This copying approach preserves routing accuracy while using minimal computational resources, as the simplified structure requires much less processing than the original trajectory data
Data Source
AI summary
A method and an apparatus for generating a road network are disclosed. The method for generating a road network comprises: aggregating a plurality of grid cells partitioned in advance on a trajectory map based on trajectories in each grid cell of the plurality of grid cells to form level-1 regions; and generating a link of the road network by merging a level-1 region having two valid neighbors with its neighbor level-1 regions having two valid neighbors.