HesaiLidar_Swift_ROS

About the project

This repository includes the ROS Driver for Pandar LiDAR sensor manufactured by Hesai Technology. Branches are included for different systems and UDP protocol versions.

Developed based on HesaiLidar_Swift_SDK, After launched, the project will monitor UDP packets from Lidar,parse data and publish point clouds frames into ROS under topic: /pandar_points. It can also be used as an official demo showing how to work with HesaiLidar_Swift_SDK.

Branches

* master: Pandar LiDAR ROS driver for Ubuntu 18.04 and Ubuntu20.04 supports the latest UDP protocol v1.4 v1.3 and v3.2 
* UDP1.4_ubuntu16.04: Pandar LiDAR ROS driver for Ubuntu 16.04 supports the latest UDP protocol v1.4
* UDP4.3: Pandar LiDAR ROS driver for  Ubuntu 16.04/Ubuntu 18.04/Ubuntu 20.04 supports the latest UDP protocol v4.3 

To get the UDP protocol version number of your device,  check the UDP package header field.

Environment and Dependencies

System environment requirement: Linux + ROS

　Recommanded:  
　Ubuntu 16.04 - with ROS kinetic desktop-full installed or  
　Ubuntu 18.04 - with ROS melodic desktop-full installed or 
  Ubuntu 20.04 - with ROS noetic desktop-full installed
　Check resources on http://ros.org for installation guide 

Compiler version requirement

 Cmake version requirement: Cmake 3.8.0 or above
 G++ version requirement: G++ 7.5.0 or above

Library Dependencies: libpcl-dev + libpcap-dev + libyaml-cpp-dev + libboost-dev

$ sudo apt-get update
$ sudo apt install libpcl-dev libpcap-dev libyaml-cpp-dev  libboost-dev

Download and Build

Install catkin_tools

$ sudo apt-get update
$ sudo apt-get install python-catkin-tools

Download code

$ mkdir -p rosworkspace/src
$ cd rosworkspace/src
$ git clone https://github.com/HesaiTechnology/HesaiLidar_Swift_ROS.git --recursive

Install required dependencies with the help of rosdep

$ cd ..
$ rosdep install -y --from-paths src --ignore-src --rosdistro $ROS_DISTRO

Build

$ catkin_make -DCMAKE_BUILD_TYPE=Release

Configuration

 $ cd src/HesaiLidar_Swift_ROS/pandar_pointcloud/launch

open PandarSwift_points.launch to set configuration parameters

Reciving data from connected LiDAR: config LiDAR IP address & UDP port, and leave the pcap_file empty

Parameter	Default Value
device_ip	192.168.1.201
port	2368
pcap_file

Data source will be read from connected LiDAR when "pcap_file" is set to empty

Reciving data from pcap file: config pcap_file and correction file path

Parameter	Value
pcap	pcap file path
calibration	lidar calibration file path

Data source will be read from pcap file instead of LiDAR once "pcap_file" not empty

Run

View the point clouds from connected LiDAR

1. Make sure current path in the rosworkspace directory

$ source devel/setup.bash
$ roslaunch pandar_pointcloud PandarSwift_points.launch

2. The driver will publish a PointCloud2 message in the topic.

/pandar_points

3. Open Rviz and add display by topic.

4. Change fixed frame to "PandarSwift" to view published point clouds.

View the point clouds from rosbag file

1. Make sure current path in the rosworkspace directory

$ source devel/setup.bash
$ roslaunch pandar_pointcloud PandarSwift_points.launch

2. The driver will publish a raw data packet message in the topic.

/pandar_packets

3. record raw data rosbag

$rosbag record -b 4096 /hesai/pandar_packets

4. stop roslaunch and rosbag record by "Ctrl + C"

5. play raw data rosbag

$rosbag play <rosbagfile>

6. launch transform_nodelet.launch

$ roslaunch pandar_pointcloud transform_nodelet.launch data_type:=rosbag

7. Open Rviz and add display by topic.
8. Change fixed frame to "PandarSwift" to view published point clouds.

Details of launch file parameters and utilities

Parameter	Default Value
calibration	Path of correction file, will be used when not able to get correction file from a connected Liar
device_ip_ip	The IP address of connected Lidar, will be used to get correction file
host_ip	The IP address of host pc, will be used to get udp data from lidar
frame_id	frame id of published messages
firetime_file	Path of firetime files
pcap	Path of the pcap file, once not empty, driver will get data from pcap file instead of a connected Lidar
port	The destination port of Lidar, driver will monitor this port to get point clouds packets from Lidar
start_angle	Driver will publish one frame point clouds data when azimuth angle step over start_angle, make sure set to within FOV
publish_model	default "points":publish point clouds "raw":publish raw UDP packets "both_point_raw":publish point clouds and UDP packets
namespace	namesapce of the launching node
coordinate_correction_flag	default "false":Disable coordinate correction "true":Enable coordinate correction
channel_config_file	Path of channel config file, will be used when not able to get channel config file from a connected Liar
cert_file	Path of the user's certificate
private_key_file	Path of the user's private key
ca_file	Path of the root certificate
multicast_ip	The multicast IP address of connected Lidar, will be used to get udp packets from multicast ip address