Virtual Lane Generation from Objects When Lane Markings Fail
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing autonomous driving technologies face challenges in accurately detecting lane boundaries, especially under adverse weather conditions or when road markings are unclear, leading to unreliable lane detection information.
Innovation Solution
A virtual lane generating method and device that uses GPS and LiDAR data to create a virtual lane based on objects detected in the external image, such as other vehicles, by clustering and estimating lane regions and generating a virtual boundary line, even when traditional lane detection information is invalid.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If traditional lane detection methods are used to detect lane boundaries from forward-view images, then the system can identify lane markings under clear conditions, but the detection reliability deteriorates under adverse weather conditions (snow, rain, fog) or when road markings are unclear
Solution Approach 1:
The patent introduces virtual lane boundary lines as an intermediary solution when traditional lane detection fails. These virtual boundaries are generated by detecting objects (vehicles, pedestrians) and inferring lane positions based on object locations and road geometry, rather than directly detecting physical lane markings. This mediator approach allows the system to maintain lane detection capability under adverse conditions where traditional methods fail.
Solution Approach 2:
The system creates virtual copies of lane boundary information by generating synthetic lane boundary lines based on detected objects and road characteristics. Instead of relying on physical lane markings, the system synthesizes lane boundary representations from alternative data sources (object positions, GPS coordinates, LiDAR data), effectively copying the essential information needed for lane keeping without requiring the original physical markings to be visible.
2Reliability
If the system generates virtual lane boundaries using object-based clustering and LiDAR data, then lane detection reliability improves under adverse conditions, but the device complexity increases due to multiple sensors and processing algorithms
Solution Approach 1:
The patent makes the object detection system serve multiple functions: it detects objects for collision avoidance purposes and simultaneously uses the same object positions to infer lane boundary locations. The LiDAR sensor performs both obstacle detection and provides depth information for virtual lane generation. This multi-functionality approach allows the system to generate virtual lanes without adding dedicated sensors solely for lane detection, thereby managing complexity while improving reliability.
Solution Approach 2:
The system merges multiple data sources (camera images, LiDAR point clouds, GPS coordinates, object detection results) into a unified virtual lane generation process. By combining these diverse inputs and processing them through integrated algorithms, the system creates a cohesive lane boundary representation that leverages the strengths of each sensor type while managing overall system complexity through unified processing architecture.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
This solution provides a reliable virtual lane for autonomous vehicles to follow, enhancing safety and navigation accuracy even in conditions where traditional lane detection fails, by using object-based clustering and LiDAR data to generate a virtual lane boundary.
Implementation Method 1
The processor may be configured to generate the virtual lane based on the object, location information collected from a Global Positioning System (GPS), and Light Detection and Ranging data (LiDAR).
Data Source
Figure 1
Figure 2
Figure 3
AI summary
A virtual lane generating method and device are provided. The virtual lane generating method includes determining validity of lane detection information extracted from an image in front a vehicle, and generating a virtual lane based on an object included in the image, in response to a determination that the lane detection information is not valid.