Geometrical Stem Detection

Description

The program allows to identify, in real-time, the stems of sugarbeets using a top-down image.

The idea is to first detect the leaves of the plant using a geometrical approach. The stem is then derived out of the intersection of all leaf directions.

Key contributors

• Ferdinand Langer (ferdinand.langer@uni-bonn.de)
• Leonard Mandtler (mandtler@uni-bonn.de)

Related publications

TODO

Dependencies

• catkin
• boost >= 1.54
• OpenCV >= 3.0
• yaml-cpp
• ROS Kinetic

Installation guide

On Ubuntu 16.04, most of the dependencies can be installed from the package manager:

sudo apt install git libopencv-dev catkin

You can install yaml-cpp by following the instructions at yaml-cpp

Additionally, make sure you have catkin-tools installed:

sudo apt install python-pip
sudo pip install catkin_tools

If you do not have a catkin workspace already, create one:

cd
mkdir catkin_ws
cd catkin_ws
mkdir src
catkin init

Clone the repository in your catkin workspace:

cd ~/catkin_ws/src
git clone https://github.com/Photogrammetry-Robotics-Bonn/geometrical_stem_detection.git

Then, build the ROS nodes for vegetation segmentation and stem detection:

catkin build piros_vegsec piros_stemdet

How to use it

This whole project is meant to run on a mobile ROS platform performing the stem detection.

An example bag file is provided in the piros_stemdet/example directory. To run the program on the bag run:

roslaunch piros_stemdet example.launch

For each image, all the detected stems are published on the topic /piros_stemdet/stem and printed on the terminal. For a visual representation, refer to the topic /piros_stemdet/mask.

For further information about the single nodes, refer to piros_vegsec README and piros_stemdet README.

License

This project is licensed under the FreeBSD License. See the LICENSE.txt file for details.