Implementation of an Autonomous Driving System for FPT2018 FPGA Design Competition Using the Zynqberry Processing Board
Yohei Shimmyo, Maiko Arakawa, Shunsuke Mie, Hiroaki Saito, Yuichi Okuyama, Hiroki Yomogita · 2018
We propose an autonomous driving system with vision-based algorithms on FPGA. Our car is composed of Zynqberry processing board, general USB Web camera, and gearbox and motors that drive facing two wheels. We built customized Ubuntu Linux including libraries for Web camera, Wi-Fi, OpenCV on ARM cores on the Zynq processor. We implement the motor control module as hardware logic on the FPGA part and connected to the Linux on processing system via AXI-bus. On the processing system, software self-localization, and path planning modules are running with the help of Linux. We tested all of the functions of our driving system. However, further software tuning is needed to control the vehicle accurately.