3D mapping with Unitree-LIDAR

3D mapping with Unitree 4D LIDAR using ROS, Ubuntu & Point cloud

The Robotics way of looking at the world!

The project aims to employ Unitree 4D LiDAR technology within the Robot Operating System on Ubuntu, facilitating a precise 3D mapping endeavor of a segment within the University of Bridgeport's Engineering building. Utilizing point cloud data and Simultaneous Localization and Mapping (SLAM) techniques, the initiative aims to create an accurate representation of the environment. This undertaking merges cutting-edge LiDAR capabilities with the efficiency of ROS on Ubuntu, enhancing the mapping process for the specified region. The project aligns with advancements in spatial technology, contributing to comprehensive understanding and documentation of the University's engineering infrastructure.

This adapts the state-of-the-art lidar inertial odometry algorithm, POINT-LIO, for use with our lidar, Unitree LIDAR L1.

POINT-LIO is a robust and high-bandwidth lidar inertial odometry (LIO) with the capability to provide accurate, high-frequency odometry and reliable mapping under severe vibrations and aggressive motions.

3D mapping .mp4

3D LiDAR Mapping.mp4

Prerequisites

Ubuntu and ROS

We tested our code on Ubuntu20.04 with ROS noetic. Ubuntu18.04 and lower versions have problems of environments to support the Point-LIO, try to avoid using Point-LIO in those systems.

You can refer to the official website to install ROS noetic:

http://wiki.ros.org/noetic/Installation/Ubuntu

Additional ROS package is required:

sudo apt-get install ros-xxx-pcl-conversions

Eigen

Following the official Eigen installation, or directly install Eigen by:

sudo apt-get install libeigen3-dev

unitree_lidar_ros

You should download and build unitree_lidar_ros follwing these steps:

git clone https://github.com/unitreerobotics/unilidar_sdk.git

cd unilidar_sdk/unitree_lidar_ros

catkin_make

Build

Clone this repository and run catkin_make:

mkdir -p catkin_point_lio_unilidar/src

cd catkin_point_lio_unilidar/src

git clone https://github.com/unitreerobotics/point_lio_unilidar.git

cd ..

catkin_make

Run

Run with Unilidar

Firstly, you should connect our lidar to your PC serial port, and supply power for the lidar with a 12V charger.

Besides, to ensure proper initialization of the IMU, it is advisable to keep the lidar in a stationary state during the initial few seconds of algorithm execution.

Run unilidar:

cd unilidar_sdk/unitree_lidar_ros

source devel/setup.bash

roslaunch unitree_lidar_ros run_without_rviz.launch

Run Point-LIO:

cd catkin_unilidar_point_lio

source devel/setup.bash

roslaunch point_lio_unilidar mapping_unilidar.launch

After completion of the run, all cached pointcloud map will be saved to the following path:

catkin_point_lio_unilidar/src/point_lio_unilidar/PCD/scans.pcd

You can use the pcl_viewer tool to view this pcd file:

pcl_viewer scans.pcd

P.S - Skipping the wait for a robot, we dove into testing the LIDAR by manually dragging a laptop through each lab. The result? A stunning 3D map of the entire space. Major shoutout to Ian Spetnagel for the crucial role in this project. His help was key to pulling this off.

Page updated

Google Sites

Report abuse