Extracting Cognition from a Front Facing Camera on Autonomous Vehicles

Abstract

Autonomous vehicles are widely seen as a solution to many of the issues drivers encounter today. Traffic, accidents, and driver distraction are all things that autonomous vehicles could minimize. Today’s autonomous vehicles usually carry very expensive Radar, LIDAR, and GPS systems to calculate speeds and distances, which can cost thousands of dollars. Those expensive features keep autonomous vehicles from being accessible to the average person. This thesis looks at the calculation and measurement of important parameters such as relative speeds, and distances between vehicles based on video supplied from a cheap front-facing camera(s) system, with the goal of having it run in real time.