Strawberry-field

About this work

This work is an example of software-in-the-loop (SIL), where trajectory planning is performed within a strawberry field using a visual memory. The process starts with a camera that compares a reference image with a desired image (the latter are stored in the visual memory).

First, ORB points are detected in both images, and then a point correspondence is established to gather information and compute the homography matrix. This matrix provides data on rotation and translation. Finally, this information is used to integrate it into the IBVS control, as presented in [1].

---

How to Run the Program

Prerequisites

You need to install OpenCV first. You can install it with the following command:

sudo apt-get update
sudo apt-get install libopencv-dev

You need to install the TurtleBot3 packages in order to use the TurtleBot in Gazebo:

• TurtleBot3 Quick Start Guide
• TurtleBot3 Simulation Setup

You also need to install the ViSP library:

• ViSP Installation Guide for Ubuntu

Additionally, install the ViSP ROS package:

• ViSP ROS GitHub Repository

Setup Instructions

1. Download the source code and place it inside a ROS workspace.

2. Compile the workspace by running:

   catkin_make

3. Modify the image reference path:

   • Open homography_vision.cpp.
   • Change the path on line 154 to the full path where the image_reference folder is located.

4. Run the program with the following commands:

   source devel/setup.bash
   export TURTLEBOT3_MODEL=turtlebot3_waffle1
   roslaunch turtlebot3_gazebo turtlebot3_fresa.launch

(Optional) Visualization

If you want to visualize the point correspondences and the desired image transformation:

• Open a new terminal in the same workspace directory.

• Run the following command:

  rqt_image_view

• Look for the topic /camera/rgb/vision_image_matches in the list.

Note: The models used for the strawberry field environment were added to
turtlebot3_simulations/turtlebot3_gazebo/models.

This program was developed and tested in ROS 1 Noetic. 🚀

---

Contributors

• @erandivg
• @gfloresc
• @NoePity2

---

References

[1] Chaumette, F., & Hutchinson, S. (2006). Visual Servo Control, Part I: Basic Approaches. IEEE Robotics & Automation Magazine, 13(4), 82–90. Link

[2] Implementación De Control Visual Para Planificación De Trayectorias En Un Cultivo De Fresas Virtual. Jóvenes en la Ciencia. Universidad de Guanajuato. Link

[3] Visual Control Based Trajectory Design for a Mobile Robot in an Agricultural Environment. Master’s Thesis. Centro de Investigaciones en Óptica, A.C. (CIO), 2025. Link