O-Net LiDAR ROS Driver

Build within Docker container

• Start Docker container: /opt/share/ros/start_ros_docker.sh $(pwd)
• Inside the Docker container, cd into the repo directory and execute

source /opt/ros/melodic/setup.bash
catkin_make

Build without Docker container

• cd into the repo directory and execute

source /opt/ros/melodic/setup.bash
catkin_make

Launch the ROS node

• After catkin_make, set environment using source devel/setup.bash and execute roslaunch lidar_ros_driver run.launch
• If save_bag set to True, the saved bag file can be found in devel/lib/lidar_ros_driver/test.bag

Create install package

• Execute catkin_make install and package the install directory

PointCloud2 to PCL PointCloud Conversion

Convert sensor_msgs::PointCloud2 objects coming from the velodyne_pointcloud driver to pcl::PointCloud<onet::lidar_ros::PointXYZIRGBT> objects.

The sensor_msgs::PointCloud2 ROS message is constructed by using the onet::lidar_ros::PointXYZIRGBT container. Both define the following channels:

• x - The x coord in Cartesian coordinates
• y - The y coord in Cartesian coordinates
• z - The z coord in Cartesian coordinates
• intensity - The measured intensity at the point
• r - RGB information at the point (calculated from intensity)
• g - RGB information at the point (calculated from intensity)
• b - RGB information at the point (calculated from intensity)
• utc_time - The time stamp (UTC) of the measured point
• ms_time - The time stamp (millisecond) of the measured point

To convert a sensor_msgs::PointCloud2 published by the velodyne driver to PCL you could use the following code:

#include <pcl_conversions/pcl_conversions.h>
#include <lidar_ros_driver/point_types.h>
#include <pcl/PCLPointCloud2.h>
#include <pcl/conversions.h>

void cloud_callback(const boost::shared_ptr<const sensor_msgs::PointCloud2>& cloud){

  pcl::PCLPointCloud2 pcl_pointcloud2;
  pcl_conversions::toPCL(*cloud, pcl_pointcloud2);
  pcl::PointCloud<onet::lidar_ros::PointXYZIRGBT>::Ptr pcl_cloud_ptr(new pcl::PointCloud<onet::lidar_ros::PointXYZIRGBT>);
  pcl::fromPCLPointCloud2(pcl_pointcloud2, *pcl_cloud_ptr);

  // use pcl_cloud_ptr
}

How to use time stamp of each point

• utc_time - The time stamp (UTC) of the measured point
• ms_time - The time stamp (millisecond) of the measured point

#include <boost/date_time/gregorian/gregorian.hpp>
#include <ctime>

boost::gregorian::date now_date = boost::gregorian::day_clock::universal_day();
        struct tm now_tm = boost::gregorian::to_tm(now_date);
        now_tm.tm_hour = (point.utc >> 12) + 8;
        now_tm.tm_min = (point.utc & 0xFC0) >> 6;
        now_tm.tm_sec = point.utc & 0x3F;
        double now_sec = (mktime(&now_tm) * 1000000 + (point.time_stamp >> 10) * 1000 +
                        (point.time_stamp & 0x3FF)) /
                        1000000.0;

支持启动多台雷达

1.启动单台

launch文件示例见./src/launch/single.launch

2.启动多台

launch文件示例见./src/launch/multiple.launch

3.多台launch文件修改 如果新增一台雷达，可以在launch文件中添加一个group，ns取值自定义，最好统一，如添加第10台雷达，修改如下：

group ns修改，ns="lidar10"

point_cloud_topic_name修改，default="lidar_point_cloud10"

frame_id修改，default="lidar10

param_path修改, default="$(find lidar_ros_driver)/config10"，在src目录下添加配置参数目录config10