Lane Detection Using HD Map and Camera Fusion
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional lane departure warning and lane keeping assist systems face challenges in accurately detecting driving lane information, especially under adverse weather conditions or when lane markings are absent, leading to reduced system performance.
Innovation Solution
A method and apparatus that combine high definition map data with front-view image data from a vehicle, using coordinate system conversion to obtain precise driving lane information, which is then used to enhance the accuracy of lane detection and vehicle control systems.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If lane information is obtained only from front view images, then the system structure remains simple, but detection accuracy deteriorates under adverse weather conditions or when lane markings are absent
Solution Approach 1:
The patent combines multiple information sources (front view images from cameras and high definition map data) to obtain lane information. By merging these different data sources and converting them to a unified coordinate system, the system achieves more reliable lane detection under adverse conditions while maintaining reasonable system complexity through integrated processing.
2Reliability
If multiple data sources are integrated to improve detection accuracy, then lane information reliability improves, but data processing complexity increases
Solution Approach 1:
The patent introduces a coordinate system conversion unit as an intermediary component that transforms map data into the vehicle's coordinate system. This mediator enables seamless integration of map information with camera images, allowing multiple data sources to be processed together without creating excessive complexity in the overall system architecture.
Solution Approach 2:
The system processes multiple types of data (image data and map data) through a unified coordinate conversion and fusion framework. This multi-functional approach allows the same processing pipeline to handle different data sources, reducing overall system complexity while improving reliability through diverse information integration.
3Measurement precision
If high definition map data is combined with image data, then detection precision improves in challenging conditions, but information processing time increases
Solution Approach 1:
The system performs coordinate system conversion on map data in advance before combining it with real-time image data. This preliminary processing of map information allows for faster integration during actual lane detection, reducing the time penalty associated with processing multiple data sources while maintaining high detection precision.
Data Source
AI summary
In accordance with an aspect of the present disclosure, there is provided a method for detecting a lane information comprising: obtaining, from a high definition map, first driving lane information corresponding to estimated position information on a vehicle; obtaining second driving lane information from a front view image from the vehicle captured by a camera installed in the vehicle; converting the first driving lane information and the second driving lane information according to an identical coordinate system; and obtaining final driving lane information by combining the converted first driving lane information and the converted second driving lane information.