Image Processing in Autonomous Vehicle Model Positioning and Movement Control

This paper presents a novel technique for processing an on-board camera's images with the goal to determine the location and trajectory of an unmanned autonomous vehicle (AV) in order to control traffic through an intersection. The empirical study with multiple AV's physical models was conducted. The physical testing ground parameters and components parts used to build up the AV's model and to develop methods for image processing and decision-making are described. The technique has been successfully employed and tested under laboratory conditions.