realsense_pointnet_on_ROS

PointNet with ROS Integration

This project demonstrates the integration of PointNet with ROS to process point clouds obtained from a Realsense D455 camera. The PointNet model is used to segment potential movable objects such as humans, robots, and chairs.

Features

• Implementation of PointNet using PyTorch
• Integration with ROS
• Processing of point clouds from Realsense D455 camera

Requirements

• Python 3.7 or later
• PyTorch
• ROS Noetic
• Realsense ROS package
• OpenCV
• NumPy

Code Structure

• src/
  • pointnet.py: Contains the implementation of the PointNet model.
  • ros_pointnet.py: Contains the ROS node for integrating PointNet.
  • mvs.py: Contains the MVS algorithm implementation for generating point clouds.
• launch/
  • realsense_pointnet.launch: Launch file for starting the ROS nodes.
• scripts/
  • mvs_node.py: ROS node for running the MVS algorithm.
  • pointnet_node.py: ROS node for running PointNet.

Setup

1. Install the required libraries:

   pip install -r requirements.txt
   

2. Clone this repository:

   git clone https://github.com/yourusername/ros_pointnet.git
   cd ros_pointnet
   

3. Build the ROS package:

   catkin_make
   source devel/setup.bash
   

Usage

To launch the ROS nodes:

roslaunch ros_pointnet realsense_pointnet.launch