Virtual Lane Detection for Vulnerable Road Users
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional automated vehicles are unable to detect vulnerable road users (VRUs) outside of designated lanes with sufficient reliability, posing a risk of collisions, especially in areas where vehicles can move or drive.
Innovation Solution
The method involves defining virtual lanes and safety areas outside the detected lanes, comparing the position of VRUs with these virtual extensions, and tracking their positions to assess collision risks, utilizing existing camera and radar sensor data without additional hardware.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If conventional lane detection methods are used, then lane markings can be detected with high reliability, but vulnerable road users outside detected lanes cannot be detected reliably
Solution Approach 1:
The patent extends the detection framework from traditional 2D lane marking detection to a 3D spatial safety zone approach. By defining virtual lanes and safety areas that extend beyond detected physical lane markings, the system captures VRUs in previously undetected regions while maintaining the same camera-based detection infrastructure, thus resolving the contradiction between detection reliability and area coverage.
Solution Approach 2:
The patent introduces virtual lanes and safety areas as intermediary structures between the detected physical lanes and the potential VRU detection zones. These virtual extensions act as mediators that bridge the gap between reliable lane detection and the need to detect VRUs in adjacent areas, enabling the system to monitor regions beyond traditional lane boundaries without requiring additional hardware.
2Adaptability or versatility
If detection area is expanded beyond lanes, then more VRUs can be detected, but computing complexity increases
Solution Approach 1:
The patent segments the detection space into distinct virtual lanes and safety areas based on the detected physical lanes. By dividing the expanded detection area into structured zones that mirror the existing lane configuration, the system maintains manageable computing complexity while achieving broader coverage. Each virtual lane and safety area can be processed independently using the same algorithms applied to detected lanes.
Solution Approach 2:
The patent makes the existing lane detection algorithm universal by applying it to both detected physical lanes and virtual extended lanes. The same processing pipeline, camera data interpretation, and tracking methods used for detected lanes are reused for virtual lanes and safety areas, avoiding the need for separate complex detection systems and thereby controlling computing complexity while expanding detection area.
3Reliability
If additional sensors are added to detect VRUs outside lanes, then detection reliability improves, but hardware cost and system complexity increase
Solution Approach 1:
The patent creates virtual copies of lane structures (virtual lanes and safety areas) that replicate the detection framework of physical lanes. By copying the existing lane detection methodology and applying it to extended regions, the system achieves reliable VRU detection outside traditional lanes without adding physical sensors or hardware components, thus maintaining hardware simplicity while improving detection reliability.
Solution Approach 2:
The system uses its existing camera-based lane detection capability to serve the additional function of detecting VRUs in virtual lanes and safety areas. The same camera data processing infrastructure that detects physical lanes is made to self-service by automatically generating virtual lane representations and safety zones, eliminating the need for additional sensors or hardware while improving overall detection reliability.
Data Source
AI summary
A method for detecting vulnerable road users by using a vehicle includes ascertaining a current lane in which the vehicle is moving, a left-hand lane that is located to the left next to the current lane and/or a right-hand lane that is located to the right next to the current lane. The method also includes defining a virtual left-hand side lane that is located to the left next to the left-hand lane and/or a virtual right-hand side lane that is located to the right next to the right-hand lane, and comparing an ascertained position of a vulnerable road user with the virtual left-hand side lane and/or the virtual right-hand side lane. The method also includes tracking the ascertained position of the vulnerable road user when the comparison reveals that the position of the vulnerable road user is in the virtual left-hand side lane and/or in the virtual right-hand side lane.