Autonomous Driving Map Generation Using Reference Mark Correction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing map generation systems for autonomous driving face challenges in ensuring accuracy, particularly when the number of vehicles uploading map information is insufficient or when GPS reception is poor, leading to potential inaccuracies in map data.
Innovation Solution
A map generation system that utilizes a reference mark with accurately measured absolute coordinates to correct the coordinates of other features, enabling reduced variation in landmark positions with less probe data and allowing for accurate map data generation and distribution based on accuracy levels, thereby restricting the use of map data for advanced driving applications.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Quantity of substance
If the number of vehicles uploading map information is insufficient, then the amount of probe data is reduced, but the accuracy of map data deteriorates
Solution Approach 1:
The patent introduces reference marks with known absolute coordinates as intermediaries between the observed features and the map coordinate system. These reference marks serve as mediators that enable accurate coordinate transformation and correction, allowing high-accuracy map generation even with limited probe data from few vehicles.
2Device complexity
If coordinate correction is not applied, then the processing is simpler, but the accuracy of map data deteriorates
Solution Approach 1:
The patent introduces reference marks with known absolute coordinates as intermediaries between the observed features and the map coordinate system. These reference marks serve as mediators that enable accurate coordinate transformation and correction, allowing high-accuracy map generation even with limited probe data from few vehicles.
3Adaptability or versatility
If all map data is used for autonomous driving, then the versatility is improved, but the safety deteriorates due to low-accuracy data
Solution Approach 1:
The patent applies local quality by assigning different accuracy levels to different regions or features in the map data. Based on the statistical analysis and coordinate correction results, the system identifies high-accuracy regions suitable for autonomous driving and separates them from low-accuracy regions, ensuring that autonomous driving functions only use verified high-accuracy map data.
Data Source
AI summary
A map generation system provides to: acquire an image of the vehicle's environment from the imaging device; analyze the image to calculate the position of a feature on the road; upload map information including the position of the feature to a server; and statistically determines the coordinates of individual features in the uploaded map information. The features include a reference mark whose absolute coordinates are determined. The coordinates of the features other than the reference mark are corrected so as to match the observation coordinates of the reference mark corresponding to the absolute coordinates of the reference mark.


