Vehicle Localization Using Sparse Maps and Lane Measurements
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in navigating efficiently due to the sheer volume of data required for traditional mapping technologies, which can limit their ability to process and store information effectively, especially when relying on detailed maps for navigation.
Innovation Solution
The development of systems and methods that utilize cameras to create and navigate with sparse maps, which are optimized for data storage and transmission, allowing for crowdsourced data collection and distribution, and utilize image analysis to identify road features and landmarks for navigation, reducing the need for excessive data storage and transfer.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional mapping technology is used for autonomous vehicle navigation, then navigation accuracy is improved, but data storage requirements and processing complexity increase significantly
Solution Approach 1:
The patent extracts only the essential navigation elements from complete maps, creating sparse maps that contain only critical features needed for vehicle localization and navigation. This selective extraction reduces data storage requirements while maintaining sufficient navigation accuracy by focusing on key landmarks, road features, and positional information rather than storing complete detailed maps.
Solution Approach 2:
The patent segments the navigation data into discrete, manageable components including vehicle position, orientation, speed, acceleration, and sparse map features. By dividing the navigation system into modular segments that process specific types of information independently, the system reduces overall processing complexity and data storage needs while maintaining navigation precision.
2Reliability
If detailed maps are stored and updated for autonomous navigation, then navigation reliability is improved, but data transmission and processing complexity increase
Solution Approach 1:
The patent extracts only the necessary navigation parameters and sparse map features required for reliable autonomous navigation, eliminating redundant data. This extraction approach maintains navigation reliability by preserving critical information while reducing data transmission volumes and processing complexity through focused data selection.
Solution Approach 2:
Instead of starting with complete detailed maps and removing unnecessary elements, the patent inverts the approach by building navigation systems from sparse, essential data elements upward. This inversion reduces complexity by design, transmitting only necessary navigation data while maintaining reliability through careful selection of critical navigation parameters and features.
3Measurement precision
If complete map data is used for navigation, then localization accuracy is improved, but computational resources and processing time increase
Solution Approach 1:
The patent extracts only the essential features needed for accurate vehicle localization from complete map data, creating sparse representations that maintain localization precision. By focusing on key landmarks, road features, and positional markers rather than processing complete detailed maps, the system reduces computational resource consumption while preserving localization accuracy.
Solution Approach 2:
The patent applies partial action by using only the necessary portion of map data required for accurate localization rather than processing complete maps. This selective approach processes sufficient information for precise vehicle positioning while avoiding the excessive computational resources that would be required to analyze and process full detailed map datasets.
Data Source
AI summary
Systems and methods are provided for vehicle localization using lane measurements. In one implementation, a method of correcting a position of a vehicle navigating a road segment may include determining, based on an output of at least one navigational sensor, a measured position of the vehicle along a predetermined road model trajectory; receiving at least one image representative of an environment of the vehicle, analyzing the at least one image to identify at least one lane marking, determining, based on the at least one image, a distance from the vehicle to the at least one lane marking, determining an estimated offset of the vehicle from the predetermined road model trajectory based on the measured position of the vehicle and the determined distance, and determining an autonomous steering action for the vehicle based on the estimated offset to correct the position of the vehicle.