Map Data Generation Using Edge-Relative Landmark Association
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in generating accurate map data due to sensor noise, GPS noise, and occlusions, which affect the localization, navigation, and path planning processes.
Innovation Solution
A system that transforms spatial coordinates of landmark locations to edge-relative coordinates, using a Global Nearest Neighbor (GNN) algorithm to improve data association, clarifying spatial relationships among landmarks and resolving ambiguities in their actual locations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If multiple vehicles capture data to generate map data, then the coverage and quantity of map data improve, but the accuracy deteriorates due to sensor noise, GPS noise, and occlusions
Solution Approach 1:
The patent segments the map data generation process into distinct modules: data collection from multiple vehicles, data association using GNN algorithm, coordinate transformation to edge-relative coordinates, and map generation. This segmentation allows each module to be optimized independently, particularly the data association module that filters noisy data from multiple sources before generating the final map.
Solution Approach 2:
The patent introduces an intermediary processing layer between raw sensor data and final map generation. The Global Nearest Neighbor algorithm acts as a mediator to associate landmark detections from multiple vehicles, and the edge-relative coordinate system serves as an intermediary representation that resolves ambiguities before final map output, thereby improving accuracy while maintaining quantity.
2Device complexity
If traditional coordinate systems are used for landmark locations, then the data processing is simpler, but the data association accuracy deteriorates due to ambiguities in landmark identification
Solution Approach 1:
The patent transforms the coordinate system from a standard global coordinate system to an edge-relative coordinate system. This dimensional transformation represents landmark positions relative to road edges rather than absolute coordinates, adding contextual information about the landmark's relationship to the road network. This resolves ambiguities in data association by providing additional spatial context, improving accuracy while the transformation itself is computationally straightforward.
Data Source
AI summary
Systems and methods for generating map data are disclosed herein. One embodiment of a map-data generation system receives, from one or more vehicles that traveled within a region, a set of estimated locations for each landmark in a plurality of landmarks within the region. The system also generates a base zone map of the region that represents roadways as edges and intersections as junctions. The system also transforms, to edge-relative coordinates, the spatial coordinates of the sets of estimated locations. The edge-relative coordinates improve a Global Nearest Neighbor (GNN) algorithm in performing data association to generate a final estimated location for each landmark. The system also outputs a final zone map that includes the final estimated location for at least one landmark. The final zone map is used for one or more of localization, navigation, and path planning to control an autonomous vehicle.