Map Information Correction Using Feature Point Priority
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional map information correction systems for dynamic maps are inefficient in updating and correcting short-period maps, leading to errors that hinder self-driving vehicle navigation and automatic driving applications.
Innovation Solution
A map information correction apparatus that prioritizes feature points based on reliability, calculates errors, and corrects dynamic map information using a feature point priority determination unit, a feature pair priority determination unit, and an error calculation unit, allowing for faster and more accurate updates.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If conventional map information correction systems are used to update dynamic maps, then map information can be corrected, but the correction process takes too much time and cannot meet the real-time requirements of automatic driving
Solution Approach 1:
The patent segments the map correction process by dividing feature points into multiple priority levels (first priority, second priority, third priority). This segmentation allows the system to process only the most critical feature points first, significantly reducing the time required for essential corrections while maintaining accuracy for less critical features in subsequent processing stages.
Solution Approach 2:
The patent implements preliminary action by pre-calculating and storing priority levels for different feature points before the actual correction process. This pre-established priority structure enables the system to immediately begin correcting high-priority features without delay, achieving real-time performance requirements for automatic driving applications.
2Measurement precision
If all feature points are corrected with equal priority, then comprehensive accuracy is maintained, but the correction process becomes too slow for dynamic map applications
Solution Approach 1:
The patent applies local quality by assigning different priority levels to different feature points based on their specific characteristics and importance. High-priority feature points (such as those critical for navigation safety) receive immediate correction with highest accuracy, while lower-priority points are processed subsequently. This differential approach maintains necessary accuracy for critical features while significantly improving overall correction efficiency.
Solution Approach 2:
The patent implements partial action by correcting only the most essential feature points (first priority) within the real-time constraint, accepting that less critical features (third priority and below) will be corrected in subsequent cycles. This partial correction approach ensures the minimum necessary accuracy for safe automatic driving operation while meeting strict time requirements.
Data Source
AI summary
A map information correction apparatus includes a feature point priority determination unit that determines a priority for a feature point of own mobile object surrounding information. A feature pair priority determination unit gives, when there is one or more feature pairs, a priority to the feature pair, based on an own mobile object surrounding information priority of the feature pair and a map information priority of the feature pair. The feature pair is a pair of an own mobile object surrounding information feature point and a received map information feature point that correspond to each other. An error calculation unit calculates an error in the map information, based on a high-priority own mobile object surrounding information feature point and a high-priority map information feature point. A correction unit corrects the map information using the error.


