Agriculture Autonomous Vehicles under Prof. Ken-Yuh Hsu (許根玉) and Prof. Shean-Jen Chen (陳顯禎), NYCU 2019/9 - Present
- Implemented SLAM system via RGB-D camera to construct 3D model of the environments and do self positioning.
- Implemented pest detection and positioning system by integrating RGB-D image and YOLO neural network.
- Developed a deep reinforcement learning control system via deep deterministic policy gradient (DDPG) to achieve automatic obstacle avoidance.
- Integrated MATLAB Simulink, C/C++ and Python all together into ROS framework and built up the control system.
- Conducted obstacle avoidance simulation on Turtlebot3 in Gazebo under ROS.
- Achieved automatic obstacle avoidance in a real agriculture field with a lab-made agriculture vehicle.
- The results were reported by Common Wealth Magazine's Smile Taiwan (天下雜誌《微笑臺灣》) in 2020/12.