Inexpensive Vehicle Detection and Estimation

Ryan Summit (Southeastern Oklahoma State University), Sam Taylor (University of Illinois)

Autonomous vehicles have been a leading research topic for many years, with one of the main goals being to create a fully functioning autonomous road vehicle. This paper focuses on the costs associated with autonomous functions, with the intention of creating affordable autonomous availability. Costs can be greatly reduced through utilizition of inexpensive sensors, such as cameras. Initially, two artifical reference points will be identified on a lead vehicle. Then visual servoing techniques will be applied to differentiate vehicle detection and vehicle distance estimation. The concept of visual servoing is based upon the idea of comparing frames of reference points and determining the change in velocity and distance compared to the previous reference points. By the end of the summer, the car will perform in "stop and go" traffic, which can then be expanded to highway following in the future.