Awesome
Generative Grasping CNN (GG-CNN) Grasping with Kinova Mico
This repository contains a ROS package for running the GG-CNN grasping pipeline on a Kinova Mico arm. For the GG-CNN implementation and training, please see https://github.com/dougsm/ggcnn.
The GG-CNN is a lightweight, fully-convolutional network which predicts the quality and pose of antipodal grasps at every pixel in an input depth image. The lightweight and single-pass generative nature of GG-CNN allows for fast execution and closed-loop control, enabling accurate grasping in dynamic environments where objects are moved during the grasp attempt.
Paper
Closing the Loop for Robotic Grasping: A Real-time, Generative Grasp Synthesis Approach
Douglas Morrison, Peter Corke, Jürgen Leitner
Robotics: Science and Systems (RSS) 2018
If you use this work, please cite:
@article{morrison2018closing,
title={Closing the Loop for Robotic Grasping: A Real-time, Generative Grasp Synthesis Approach},
author={Morrison, Douglas and Corke, Peter and Leitner, Jürgen},
booktitle={Robotics: Science and Systems (RSS)},
year={2018}
}
Installation
This code was developed with Python 2.7 on Ubuntu 16.04 with ROS Kinetic. Python requirements can be found in requirements.txt
.
You will also require the Kinova ROS Packages and Realsense Camera Packages.
A 3D printed mount for the Intel Realsense SR300 on the Kinova Mico arm can be found in the cad
folder.
GG-CNN Model
See https://github.com/dougsm/ggcnn for instructions for downloading or training the GG-CNN model.
Running
This implementation is specific to a Kinova Mico robot and Intel Realsense SR300 camera.
Once the ROS package is compiled and sourced:
- Lanuch the robot
roslaunch kinova_bringup kinova_robot.launch kinova_robotType:=m1n6s200
- Start the camera
roslaunch ggcnn_kinova_grasping wrist_camera.launch
- Run the GG-CNN node
rosrun ggcnn_kinova_grasping run_ggcnn.py
- To perform open-loop grasping, run
rosrun ggcnn_kinova_grasping kinova_open_loop_grasp.py
, or to perform closed-loop grasping runrosrun kinova_closed_loop_grasp.py
.
Contact
Any questions or comments contact Doug Morrison.