Target Vehicle Lane Positioning with Map-Corrected Lateral Offset
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current systems face challenges in accurately determining the location of target vehicles relative to lane boundaries, particularly due to longitudinal errors in perception-based techniques and inaccuracies in map data, which can lead to inconsistent and unreliable navigation in autonomous driving systems.
Innovation Solution
The method involves obtaining images of target vehicles and lane boundaries, determining the distance between them, and adjusting the vehicle's position in a map based on this distance and lane boundary position, using a combination of image-based and sensor data fusion, such as RADAR and LIDAR, to ensure accurate lateral offset calculations and consistent mapping.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If perception-based techniques are used to determine vehicle location relative to lane boundaries, then the system can operate without HD maps, but longitudinal errors cause inaccurate location determination
Solution Approach 1:
The patent combines multiple data sources including perception-based lane boundary detection, vehicle sensor data (RADAR, LIDAR, IMU), and map data into a unified framework. This fusion allows the system to leverage the adaptability of perception-based methods while compensating for their longitudinal errors through complementary sensor inputs and data fusion algorithms.
2Measurement precision
If map data is used to correct perception errors, then location accuracy improves, but map data inaccuracies introduce new errors
Solution Approach 1:
The system implements a feedback mechanism where the vehicle's actual position, determined through sensor fusion and perception, is continuously compared with map-based expected positions. Discrepancies are used to adjust and correct both perception and map data, creating a self-correcting system that maintains accuracy even when individual data sources contain errors.
Solution Approach 2:
The patent creates a composite positioning system that integrates multiple data sources with different error characteristics. By combining perception data (good at lateral accuracy, poor at longitudinal), sensor data, and map data (which may have its own inaccuracies), the system achieves robust and reliable location determination that is resilient to errors in any single source.
3Reliability
If multiple sensor types are fused to improve accuracy, then location determination becomes more robust, but system complexity increases
Solution Approach 1:
The patent segments the complex sensor fusion problem into distinct functional modules: perception modules for detecting lane boundaries and vehicles, sensor fusion modules for integrating data from multiple sources, and correction modules for adjusting positions based on discrepancies. This modular architecture manages complexity by organizing the integration of RADAR, LIDAR, IMU, and camera data into manageable, independent components that can be developed and maintained separately.
Data Source
AI summary
Systems and techniques are described herein for determining at least one location of at least one target vehicle relative to a lane. For instance, a method for determining at least one location of at least one target vehicle relative to a lane is provided. The method may include obtaining a position of a target vehicle within an image; obtaining one or more positions of a lane boundary within the image; determining a distance between the target vehicle and the lane boundary based on the position of the target vehicle within the image and the one or more positions of the lane boundary within the image; and adjusting a position of the target vehicle in a map based on the distance between the target vehicle and the lane boundary and a position of the lane boundary in the map