!! NOTE !!

This project is deprecated. I've created this project in 2017 for my master thesis. Since then, no signifiant changes have been made. Since I'm not working with ROS anymore, I will probably not be able to answer your questions. Feel free to use the code and concepts as building blocks or as an inspiration for your own project. Don't expect this project to run by "plug and play" or to even compile with current versions of ROS.

Summary

A simple "framework" for ROS-based 2D and 3D object recognition software.

Documentation

You find a description of the project in docs/project_description.pdf. docs/thesis.pdf contains a more detailed description but it's not up-to-date anymore. The following video provides an overview and demonstrations: https://www.youtube.com/watch?v=Pmur4kxrexE

Requirements

ROS OpenCV 3 (only for 2D object recognition) PCL 1.7 (only for 3D object recognition) Eigen 3 (only for 3D object recognition) Qt5 (not necessary if you do not compile the GUI-files)

Compilation

1. Follow ROS installation (Desktop Full Install): http://wiki.ros.org/kinetic/Installation/Ubuntu I needed to enable mysql (sudo systemctl enable mysql) for this to work.
2. Create catkin workspace: mkdir -p ~/catkin_ws/src cd ~/catkin_ws catkin_make
3. Clone project: cd ~ git clone https://github.com/joffman/ros_object_recognition.git
4. Copy/Move src folder into your workspace. cp -r ~/ros_object_recognition/src/* ~/catkin_ws/src/
5. Compile: cd ~/catkin_ws catkin_make

License

MIT license