Abstract
Global positioning system (GPS) receivers assist autonomous vehicles and drivers to determine their exact position along the roadway. The accuracy of GPS, however, is near 6-10 meters. This accuracy can be questioned in remote or densely packed areas like tunnels or cities with multiple obstacles and tall structures where the GPS signal is disturbed frequently. Recent advances in technology have demanded more precise positioning to better the safety of autonomous vehicles, drivers, and pedestrians. In this project, a vision-based positioning technique is proposed to complement current GPS receivers and improve their accuracy. Vision sensors provide unique spatial signatures which will be processed using image processing techniques to interpolate the position of the driver. The proposed technique is expected to enhance the accuracy of GPS receivers, especially in densely packed areas and tunnels.