Technical Paper
Assessing Resilience in Lane Detection Methods: Infrastructure-Based Sensors and Traditional Approaches for Autonomous Vehicles
2024-04-09
2024-01-2039
On-board sensors in traditional perception subsystems used in autonomous vehicles (AVs) have the limitations of high computational load and data duplication. Infrastructure-based sensors are a potential alternative to traditional AV subsystems as they address the limitations mentioned above. However, these technologies are still in the early stages of development and have not undergone extensive real-world deployment and testing for lane detection systems. Thus, there is a lack of substantial data pertaining to their resilience in these systems compared to conventional methods (camera vs. infrastructure) when encountering hazardous scenarios, such as lane line occlusion, sensor failure, and environmental obstructions, and thus their resilience. This study aimed to evaluate the influence of hazards on the resilience of three different lane detection methods.