Grasp Pose Generator

Grasp Pose Generator is a Python package that provides bindings for the C++ Grasp Pose Generator (GPG), enabling the creation of grasp candidates for 3D point clouds. This tool is essential for robotic manipulation tasks, allowing robots to determine feasible grasping positions on objects represented by point clouds.

Features

• Generate 6-DOF grasp poses for 2-finger grippers.
• Evaluate grasp candidates for antipodal properties using geometric conditions.
• Seamless integration with Python for easy incorporation into robotic pipelines.

Installation

To install the grasp_pose-generator package, use pip:

pip install grasp-pose-generator

Build

To build the grasp-pose-generator package from source, follow these steps:

Steps

1. Install Vcpkg and configure it on your system.

2. Clone the Pybind11 repository into a dependencies directory in your project root:

   mkdir dependencies
   cd dependencies
   git clone https://github.com/pybind/pybind11.git
   

3. Install the package:

   pip install .
   

This process will compile the C++ code using the Vcpkg dependencies and the Pybind11 bindings.

Usage

Here's a basic example demonstrating how to generate grasp poses for a given point cloud:

import numpy as np
import grasp_pose_generator as gpg

# Generate a random point cloud (replace with your actual point cloud data)
points = np.random.rand(3000, 3)  # n x 3 numpy array

# Parameters
num_samples = 10000
show_grasp = False
gripper_config_file = "path/to/gripper_params.cfg"  # Replace with the path to your gripper configuration file

# Generate grasps
grasps = gpg.generate_grasps(points, num_samples, show_grasp, gripper_config_file)

# Convert grasps to transformation matrices
pose_list = []
for grasp in grasps:
    pose = np.eye(4)
    pose[:3, 0] = grasp.get_grasp_approach()
    pose[:3, 1] = grasp.get_grasp_binormal()
    pose[:3, 2] = grasp.get_grasp_axis()
    pose[:3, 3] = grasp.get_grasp_bottom()
    pose_list.append(pose)

Ensure that you have a valid gripper configuration file. An example configuration file can be found in the gripper_params.cfg.

Example with a Simple Box

To visualize grasp generation on a simple box, you can run the provided example script:

python example/example.py

Press R during the visualization to see the generated grasp poses.

Build

Prerequisites

• Python 3.11 or later
• Vcpkg installed and configured
• A C++ compiler (e.g., GCC, MSVC, Clang)

License

This project is licensed under the MIT License.

Acknowledgments

This package is an improved and extended version of the original pyGPG by Hongzhuo Liang. The core grasp generation functionality is based on the Grasp Pose Generator (GPG) by Andreas ten Pas.