Target Vehicle Trajectory Matching for Sparse-Map Road Positioning
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Autonomous vehicles face challenges in navigating effectively due to the sheer volume of data required for processing and storing visual information, map data, and sensor data, which can limit their navigation capabilities and pose daunting storage and update challenges.
Innovation Solution
The use of cameras and processing devices to analyze images, determine vehicle location, and calculate navigational actions based on elevation and lane width information, along with the implementation of a sparse map system that stores polynomial representations of road features and trajectories, reducing data storage needs and enabling efficient navigation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If traditional mapping technology is used to navigate, then navigation accuracy is improved, but data storage requirements and update complexity increase significantly
Solution Approach 1:
The patent extracts only the essential navigational elements from complete map data by using sparse map representations. Instead of storing and processing full map datasets, the system extracts key trajectory points, road geometry parameters, and essential landmarks that are sufficient for navigation, thereby reducing data storage requirements while maintaining navigation accuracy.
Solution Approach 2:
The patent inverts the traditional approach by not starting with complete map data and then filtering it, but rather by directly creating sparse map representations from sensor data and trajectory information. This inversion allows the system to build navigation maps with only the necessary information from the ground up, avoiding the storage burden of complete maps.
2Reliability
If complete map data is stored and updated, then navigation reliability is improved, but system complexity and update burden increase
Solution Approach 1:
The system extracts only the critical elements needed for reliable navigation from complete map data, storing sparse representations that include essential trajectory points, road geometry, and key landmarks. This extraction maintains navigation reliability by preserving the most important navigational information while eliminating redundant data that contributes to system complexity.
3Measurement precision
If extensive sensor data is processed in real-time, then navigation accuracy is improved, but processing time and computational load increase
Solution Approach 1:
The patent extracts only the essential features from sensor data that are necessary for accurate navigation, such as key trajectory points, road boundary detections, and significant landmarks. By processing and storing only these extracted essential features rather than all raw sensor data, the system maintains high navigation accuracy while significantly reducing real-time processing time and computational load.
4Measurement precision
If detailed map information is stored, then navigational decision-making accuracy is improved, but data management burden increases
Solution Approach 1:
The system extracts and stores only the critical map information elements needed for navigational decision-making, such as trajectory points, road geometry parameters, and essential landmarks. This selective extraction maintains accurate navigational decision-making while dramatically simplifying data management operations, as the sparse map structure requires less storage capacity and fewer updates compared to detailed complete maps.
Data Source
AI summary
Systems and methods are provided for navigating a host vehicle. In an embodiment, a processing device may be configured to receive images captured over a time period; analyze images to identify a target vehicle; receive map information associated including a plurality of target trajectories; determine, based on analysis of the images, first and second estimated positions of the target vehicle within the time period; determine, based on the first and second estimated positions, a trajectory of the target vehicle over the time period; compare the determined trajectory to the plurality of target trajectories to identify a target trajectory traversed by the target vehicle; determine, based on the identified target trajectory, a position of the target vehicle; and determine a navigational action for the host vehicle based on the determined position.