P-AgBot: In-Row & Under-Canopy Agricultural Robot for Monitoring and Physical Sampling

This repository contains a code for crop monitoring modules (corn height, corn stalk radius) for ROS compatible UGVs. For corn height estimation, 3D point cloud from a vertically mounted 3D LiDAR Ouster OS1-64 are used. A 2D laser scan from a inexpensive 2D LiDAR ROBOTIS LDS-01 is used to estimate each corn stalk radius. The result of each module of our system can be found here. -> https://youtu.be/CU-CX41oVyk

Dependency

• ROS (Tested with melodic.)
• Clearpath Jackal simulation (Follow the instructions to use the simulation.)

Compile

You can use the following commands to download and compile the package.

cd ~/catkin_ws/src
git clone https://github.com/kimkt0408/P-AgBot.git
cd ..
catkin_make

The system

P-AgBot consists of a unique combination and configuration of integrated sensors. P-AgBot and its components are as follows: (1) Tracking camera, (2) 3D LiDAR sensor, (3) Two-finger style gripper, (4) RGB-D camera, (5) Six degree-of-freedom robotic arm, (6) Servo motor, (7) 3D printed linkage with nichrome wire end-effector, and (8) 2D LiDAR sensor.

Run the package

1. Run the simulation world launch file:

roslaunch jackal_simulation_iros.launch

2. Run the localization launch file:

roslaunch amcl_lds.launch

3. Run the navigation launch file:

roslaunch autonomous_navigation.launch

4. (1) Run the crop height estimation launch file:

roslaunch crop_height.launch

4. (2) Run the stalk radius estimation launch file:

roslaunch stalk_radius.launch