GPU Implementation for Automatic Lane Tracking in Self-Driving Cars

The development of efficient algorithms has been the focus of automobile engineers since self-driving cars become popular. This is due to the potential benefits we can get from self-driving cars and how they can improve safety on our roads. Despite the good promises that come with self-driving cars development, it is way behind being a perfect system because of the complexity of our environment. A self-driving car must understand its environment before it makes decisions on how to navigate, and this might be difficult because the changes in our environment is non-deterministic. With the development of computer vision, some key problems in intelligent driving have been active research areas. The advances made in the field of artificial intelligence made it possible for researchers to try solving these problems with artificial intelligence. Lane detection and tracking is one of the critical problems that need to be effectively implemented. The ability of a self-driving car to successfully drive from point A to point B without going off track is dependent on lane tracking. Lane tracking in self-driving cars is a computationally intensive task and a fast implementation is needed to help a self-driving car track lanes in real-time to make the right decision at the right time. Lane tracking in self-driving cars is also dependent on the visibility of lane markings on the road. It will be difficult for a self-driving car to track lanes if the lane marking has faded, blocked by an object, or there were no lane markings on the road. Most available lane tracking implementations in the literature do not give account to these two problems. Our implementation is to solve these two problems by using artificial intelligence techniques to track lanes in all conditions and using GPU computing on NVIDIA Jetson TX2 to speed-up the process.