Virtual Driving Lane Generation Using Nearby Vehicle Orientation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing vehicle lane detection systems face challenges in accurately detecting lanes when they are obscured by nearby vehicles or under inclement weather conditions such as snow, rain, or fog, which can lead to hazardous driving situations.
Innovation Solution
A method and apparatus that recognize nearby vehicles, extract feature information indicating their directionality, and generate a virtual driving lane based on this information, using both front-view and side-view images captured by the host vehicle, to assist in lane keeping, lane changing, and speed adjustment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If lane detection is performed using traditional image processing methods, then the system is simple to implement, but the detection accuracy deteriorates when lanes are obscured by nearby vehicles or inclement weather
Solution Approach 1:
The patent introduces nearby vehicles as intermediary objects to infer lane information. Instead of directly detecting obscured lanes, the system detects vehicles' positions and orientations, then uses these as mediators to calculate and generate virtual lane markings, thereby solving the problem of obscured lane detection
Solution Approach 2:
The patent creates virtual lane markings as copies of actual lanes by synthesizing lane information from vehicle feature data. These virtual lane copies provide accurate lane guidance even when physical lanes are obscured, maintaining detection accuracy without requiring direct visual contact with the actual lane markings
2Reliability
If multiple feature extraction methods are used to improve lane detection reliability, then the measurement precision improves, but the processing time and computational complexity increase
Solution Approach 1:
The patent performs preliminary detection and extraction of vehicle feature information (positions, orientations, dimensions) before lane generation. By pre-processing and storing vehicle data, the system reduces real-time computational burden during lane synthesis, maintaining high reliability while reducing processing time
Solution Approach 2:
The patent divides the lane detection task into separate modules: vehicle detection module, feature extraction module, and virtual lane generation module. This segmentation allows each module to specialize in specific operations, improving overall efficiency and reliability while managing computational complexity through modular processing
3Measurement precision
If virtual lane generation uses comprehensive feature information from multiple vehicles, then the accuracy of driving direction determination improves, but the device complexity increases
Solution Approach 1:
The patent applies different processing strategies to different spatial regions: vehicles in front provide traffic flow direction information, while vehicles on sides provide lane boundary information. This local quality approach optimizes feature utilization for each region, improving driving direction accuracy while managing processing complexity through region-specific algorithms
Data Source
AI summary
Disclosed is a virtual driving lane generation method and apparatus for recognizing vehicles near a host vehicle, extracting feature information indicating a directionality of the nearby vehicles, generating a virtual auxiliary driving lane based on the feature information, and generating a virtual driving lane for the host vehicle based on the virtual auxiliary driving lane and the side line information.