Velodyne ROS 节点 时间戳以及话题

1. 概述

阅读Velodyne在ROS下VLP-16雷达节点源代码，将有用的部分总结一下，主要是时间戳以及topic的处理。

2. 溯源

雷达的节点源代码使用了nodelet这个概念，先慢慢看吧。根据源代码采用倒推的方式查找时间戳定义以及数据的读取代码。
1
构造函数 Convert::Convert中进行了节点初始化，主要就是从参数服务器中读取设定值。包括最小角度，最大角度等。

/** @brief Constructor. */
  Convert::Convert(ros::NodeHandle node, ros::NodeHandle private_nh, std::string const & node_name):
    data_(new velodyne_rawdata::RawData()), first_rcfg_call(true),
    diagnostics_(node, private_nh, node_name)
  {
   
    // Get startup parameters
    private_nh.param<std::string>("fixed_frame", config_.fixed_frame, "velodyne");
    private_nh.param<std::string>("target_Frame", config_.target_frame, "velodyne");
    private_nh.param<double>("min_range", config_.min_range, 10.0);
    private_nh.param<double>("max_range", config_.max_range, 200.0);
    private_nh.param<bool>("organize_cloud", config_.organize_cloud, false);

    boost::optional<velodyne_pointcloud::Calibration> calibration = data_->setup(private_nh);
    if(calibration)
    {
   
        ROS_DEBUG_STREAM("Calibration file loaded.");
        config_.num_lasers = static_cast<uint16_t>(calibration.get().num_lasers);
    }
    else
    {
   
        ROS_ERROR_STREAM("Could not load calibration file!");
    }

    if(config_.organize_cloud)
    {
   
      container_ptr_ = boost::shared_ptr<OrganizedCloudXYZIRT>(
          new OrganizedCloudXYZIRT(config_.max_range, config_.min_range,
              config_.target_frame, config_.fixed_frame,
              config_.num_lasers, data_->scansPerPacket()));
    }
    else
    {
   
      container_ptr_ = boost::shared_ptr<PointcloudXYZIRT>(
          new PointcloudXYZIRT(config_.max_range, config_.min_range,
              config_.target_frame, config_.fixed_frame,
              data_->scansPerPacket()));
    }


    // advertise output point cloud (before subscribing to input data)
    output_ =
      node.advertise<sensor_msgs::PointCloud2>("velodyne_points", 10);

    srv_ = boost::make_shared <dynamic_reconfigure::Server<velodyne_pointcloud::
      CloudNodeConfig> > (private_nh);
    dynamic_reconfigure::Server<velodyne_pointcloud::CloudNodeConfig>::
      CallbackType f;
    f = boost::bind (&Convert::callback, this, _1, _2);
    srv_->setCallback (f);

    // subscribe to VelodyneScan packets
    velodyne_scan_ =
      node.subscribe("velodyne_packets", 10,
                     &Convert::processScan, (Convert *) this,
                     ros::TransportHints().tcpNoDelay(true));

    // Diagnostics
    diagnostics_.setHardwareID("Velodyne Convert");
    // Arbitrary frequencies since we don't know which RPM is used, and are only
    // concerned about monitoring the frequency.
    diag_min_freq_ = 2.0;
    diag_max_freq_ = 20.0;
    using namespace diagnostic_updater;
    diag_topic_.reset(new TopicDiagnostic("velodyne_points", diagnostics_,
                                       FrequencyStatusParam(&diag_min_freq_,
                                                            &diag_max_freq_,
                                                            0.1, 10),
                                       TimeStampStatusParam()));
  }

需要说明
雷达的对数在校准文件中读取，我使用的是16通道，因此后面都按照num_lasers = 16来进行

    boost::optional<velodyne_pointcloud::Calibration> calibration = data_->setup(private_nh);
    if(calibration)
    {
   
        ROS_DEBUG_STREAM("Calibration file loaded.");
        config_.num_lasers = static_cast<uint16_t>(calibration.get().num_lasers);
    }

ROS下处理过后的雷达数据topic为velodyne_points，消息类型为sensor_msgs::PointCloud2

    // advertise output point cloud (before subscribing to input data)
    output_ =
      node.advertise<sensor_msgs::PointCloud2>("velodyne_points"