This is a ROS node for using YOLOv5 for real time point cloud segmentation. It uses segmentation provided in the official YOLOv5 repository. This project is adapted from Base ROS YOLOv5 implementation
This package is built and tested on Ubuntu 20.04 LTS and ROS Noetic with Python 3.8.
- Clone the packages to ROS workspace and install requirement for YOLOv5 submodule:
cd <ros_workspace>/src
git clone https://github.com/mats-robotics/detection_msgs.git
git clone --recurse-submodules https://github.com/tharaka27/ypc_ros.git
cd ypc_ros/src/yolov5
pip3 install -r requirements.txt # install the requirements for yolov5- Build the ROS package:
cd <ros_workspace>
catkin build ypc_ros # build the ROS package- Make the Python script executable
cd <ros_workspace>/src/ypc_ros/src
chmod +x detect.py
chmod +x cluster.py
chmod +x camera_info.py
- Launch the node:
roslaunch ypc_ros ypc.launch- Put your weights into
ypc_ros/src/yolov5 - Put the yaml file for your dataset classes into
ypc_ros/src/yolov5/data - Change related ROS parameters in yolov5.launch:
weights,data
- YOLOv5 official repository: https://github.com/ultralytics/yolov5
- YOLOv3 ROS PyTorch: https://github.com/eriklindernoren/PyTorch-YOLOv3
- Darknet ROS: https://github.com/leggedrobotics/darknet_ros
- YOLOv5 ROS : https://github.com/mats-robotics/yolov5_ros