Lane Detection and Following Approach in Self-Driving Miniature Vehicles

Abstract: The current work, discusses the various experiences, lessons learned and developed novelties, regarding the lane detection and lane following features of miniature autonomous vehicles. Particularly, we outline the work that was conducted, during our team’s participation in Carolo Cup 2016, which is an international competition on autonomous vehicle, held in Braunschweig, Germany. An overview of selected works from the relevant literature is presented, in order to illustrate the various techniques that are commonly adopted in order to implement the lane detection and following functions. The findings, in this regard, constitute of a set of the mostly referenced methods. Additionally, our adopted approach on the matter is extensively elucidated. The different improvements that were introduced to the existing algorithm, are quantified and visualized through an evaluation tool that was developed for this purpose. We conclude that the use of the said tool for analysis, increases the quality of the final product, which is of significant importance in embedded system, due to the fusion of software and hardware elements.