Effects of Applying Identified Road Lane Lines on Vehicle Autopilot Model Driving Performance

Autonomous Vehicle driving is one of the most actively developed systems in the automotive industry. Producing a reliable autonomous driving system is a complex multistage problem. The road lane lines are traffic signs with high level of semantics that determine how a vehicle should move, what actions are allowed to be performed, etc. Comparing different modifications of the PilotNet neural network model, the usage of detected lane lines, and their implications on vehicle autopilot driving performance is the topic of this paper. The lane lines are identified by neural network that has been trained by processing images with previously marked lane lines from CULane dataset. Vehicle autopilot driving performance is analyzed using autopilot model running on Euro Truck Simulator 2. Proposed models are analyzed and compared in multiple driving and weather environments including highway, off-road and urban driving during day, night and rain conditions. The lane line recognition (classified by a U-Net neural network model from images) is compared with additional control and path planning to determine whether it leads to better autonomous driving performance and smaller driving system model.