Vision-based Unmanned Vehicle Platform and SLAM Implementation

Project Information

  • Sponsor: Rockwell Collins
  • Leader: Chiyu Zhang
  • Participants
    • Graduate students: Githin John
    • Undergraduate students: Jaehyeok Kim (Jefferson), Jong-Min Park, Prit Chovatiya, and more

Introduction

  • The overall goal of this project is to build a vision-based unmanned vehicle platform which integrates visual sensors and algorithms to the unmanned vehicle navigation/control system.
  • The platform will be served as an algorithm testbed for the future research.
  • Current work
    • Unmanned vehicle building/testing
    • Programming/software developing in ROS (robotic operating system)
    • Visual sensor developing

Hardware

  • Vehicles
  • Visual sensors
    • Astra camera
    • 3D LIDAR (coming soon)
  • On-board computer
    • Nvidia jetson

Jetson working with Pixhawk and Astra camera

SLAM Implementation

  • As an application of the vehicle platform, we are currently implement the SLAM algorithm.
  • SLAM software: R-TAB MAP
    • An open-source SLAM program

Lab mapping with r-tab map