1.pcl_tutorial官方给的模板   
int main(int argc, char** argv)
{
  pcl::PCLPointCloud2::Ptr cloud (new pcl::PCLPointCloud2 ());
  pcl::PCLPointCloud2::Ptr cloud_filtered (new pcl::PCLPointCloud2 ());
  //建立两个容器，容纳输入点云和输出点云，pcl格式．
  
  //Fill in the point cloud  
  pcl::PCDReader::reader;
  reader.read ("filename.pcd", *cloud);  //把pcd格式的点云作为输入点云  
  std::cerr << "pointcloud number before filtering:"<< cloud->width * cloud->height <<"data points ("<< pcl::getFieldsList (*cloud) << ")."
  //cloud->width,cloud->height 点云的宽高相乘为点云的数量．
  //pcl::getFieldsList()  用于获得*cloud点云的包含的点云信息种类　（x,y,z,intensity）
  
  //create the filtering object
  pcl::VoxelGrid<pcl::PCLPointCloud2> sor;
  sor.setInputCloud (cloud);
  sor.setLeafSize (0.01f,0.01f,0.01f);
  sor.filter (*cloud_filtered);
　
　std::cerr << "pointcloud number after filtering" << cloud_filtered->width * cloud_filtered->height << "(" << pcl::getFeildsList (*cloud_filtered) <<").";
  
  pcl::PCDWriter writer;
  writer.writer ("output_filename.pcd", *cloud_filtered ,Eigen::Vector4f::zero () , Eigen::Quaternion f::Identify (), false);
 
  return (0);
}
  编写一个点云处理程序的步骤： １．主函数 ２．编写点云输入输出容器 ３．读取文件并赋值给输入容器 ４．点云处理并赋值给输出容器 ５．输出并保存   ２．结合ros的pcl.   
#include <ros/ros.h>
// PCL specific includes
#include <sensor_msgs/PointCloud2.h>
#include <pcl_conversions/pcl_conversions.h>
#include <pcl/point_cloud.h>
#include <pcl/point_types.h>
 
ros::Publisher pub
 
void cloud_cb(const sensor_msgs::PointCloud2ConstPtr& input)
{
  //create a container for data
  sensor_msgs::PointCloud2 output;
  
  //do data processing here...
  output = *input;
  
  //publish the data
  pub.publish(output);
}
 
 
int main (int argc , char** argv)
{
  //initialize ros
  ros::init (argc, argv, "package_name");
  ros::NodeHandel nh;
 
  //create a ros subscriber for the input point cloud
  ros::Subscriber sub = nh.subscribe("input",1,cloud_cb);
  //create a ros publisher for the output point cloud
  pub = nh.advertise<sensor_msgs::PointCloud2> ("output",1);
 
  //spin
  ros::spin ();
}
  编写主函数节点 int main (int argc,char** argv) １．初始化节点　ros::init ( ) ２．实例化节点 ros::NodeHandle nh; ３．订阅话题　　ros::Subsriber sub=nh.subscribe (“input”,1,cloud_cb); ４．发布话题　　ros::Publisher pub=nh.advertise<sensor_msgs::PointCloud2> (“output”,1) ５．ros::spin ()   编写点云处理总函数 void cloud_cb (const sensor_msgs::PointCloud2ConsrPtr& input) 1.输出容器 2.点云处理（包括点云数据类型转换） 3.发布话题？？？这个不完善啊，我去看一下自己的代码．   ３．90度点云分割　(ros+rviz)   １）头文件　headfilename.h 头文件里包含　要引用的头文件；常数；结构体；类．等   
#pragma once
 
#include <ros/ros.h>
#include <pcl/point_cloud.h>
#include <pcl_conversions/pcl_conversions.h>
#include <pcl/point_types.h>
#include <pcl/conversions.h>
#include <pcl_ros/transforms.h>
 
#include <pcl/filters/extract_indices.h>
 
#include <sensor_msgs/PointCloud2.h>
#include<pcl/io/pcd_io.h>//which contains the required definitions to load and store point clouds to PCD and other file formats.
 
#define CLIP_COR 0 //     x  zuobiao
 
Class PclTestCore
{
  
private:
  ros::Subscriber sub_point_cloud_;
  ros::Publisher pub_point_cloud_;
  
  void point_cb (const sensor_msgs::PointCloud2ConstPtr &in_cloud);
  void clip_above (double CLIP_COR, const pcl::PointCloud<pcl::PointXYZI>::Ptr in,const pcl::PointCloud<pcl::PointXYZI>::Ptr out)
  void clip_xy (const pcl::PointCloud<pcl::PointXYZI>::Ptr in,const pcl::PointCloud<pcl::PointXYZI>::Ptr out)
 
public:
  PclTestCore (ros::NodeHandle &nh);
  ~PclTestCore();
  void Spin();
};
  1.头文件编写 2.常数 3.编写类 Class PclTestCore private: ros::Subscriber sub_point_cloud_;ros::Publisher pub_point_cloud_;void point_cb(const sensor_msgs::PointCloud2ConstPtr &in_cloud);void clip_above(double CLIP_COR,const pcl::PointCloud::Ptr in,const pcl::PointCloud::Ptr out) 和其他点云处理函数 public: PclTestCore(ros::NodeHandle &nh);~PclTestCore(); void Spin()  
  1. _node:主函数节点
  
#include <pcl/point_types.h>
#include <pcl/conversions.h>
#include <pcl_ros/transforms.h>
 
#include <pcl/filters/extract_indices.h>
 
#include <sensor_msgs/PointCloud2.h>
#include<pcl/io/pcd_io.h>//which contains the required definitions to load and store point clouds to PCD and other file formats.
 
#define CLIP_COR 0 //     x  zuobiao
 
int main(int argc , char **argv)
{
  ros::init(argc,argv,"package_name");
  ros::NodeHandle nh;
  PclTestCore core(nh);
  return 0;
}
  主节点编写： １．节点初始化　ros::init(argc,argv,“package_name”) ２．节点实例化　ros::NodeHandle nh ３．对节点进行操作　PclTestCore core(nh) ４．return 0   ３）core_node 节点编写   
#include "   .h"
PclTestCore::PclTestCore(ros::NodeHandel &nh)
{
  sub_point_cloud_ = nh.Subcriber("/velodyne_points",10,&PclTestCore::point_cb,this)
  in_publisher = nh.advertise<sensor_msgs::PointCloud2>("/cloud_msg",10)
 
  ros::spin();
}
 
PclTestCore::~PclTestCore(){ }
void PclTestCore::spin() { }
 
void PclTestCore::clip_above(double clip_cor,const pcl::PointCloud<pcl::PointXYZ>::Ptr in,const pcl::PointCloud<pcl::PointXYZI>::Ptr out)
{
  pcl::ExtractIndices<pcl::PointXYZI> cliper;
  cliper.SetInputCloud(in);
  pcl::PointIndices indices;
  for (size_t i=0; i<in->points.size();i++)
  {
    if(in->points[i].x >clip_cor)
    {
      indices.indices.push_back(i);
    }
  }
  cliper.setIndices(boost::make_shared<pcl::PointIndices>(indices));
  cliper.setNegative(true);
  cliper.filter(*out);
  
}
 
void PclTestCore::clip_xy (const pcl::PointCloud<pcl::PointXYZI>::Ptr in,const pcl::PointCloud<PointXYZI>::Ptr out)
{
  pcl::ExtractIndices<pcl::PointXYZI> cliper;
  pcl::setPointCloud(in);
  pcl::PointIndices indices;
  
  for(size_t i=0;i<in->points.size();i++)
  {
    if(in->points[i].x<in->points[i].y)
    {
      indices.indices.push_back(i);
    }
  }
 
  cliper.setIndices(boost::make_shared<pcl::PointIndices>(indices));
  cliper.setNegative(true);
  cliper.filter(*out);
}
 
 
ros::Publisher in_publisher;
void PclTestCore::point_cb(const sensor_msgs::PointCloud2ConstPtr &in_cloud_ptr)
{
    pcl::PointCloud<pcl::PointXYZI>::Ptr current_pc_ptr(new pcl::PointCloud<pcl::PointXYZI>);
    pcl::PointCloud<pcl::PointXYZI>::Ptr cliped_pc_ptr(new pcl::PointCloud<pcl::PointXYZI>);
 
    pcl::fromROSMsg(*in_cloud_ptr, *current_pc_ptr);
 
    clip_above(CLIP_COR, current_pc_ptr, cliped_pc_ptr);
    
    pcl::PointCloud<pcl::PointXYZI>::Ptr remove_close(new pcl::PointCloud<pcl::PointXYZI>);
 
    clip_xy(cliped_pc_ptr, remove_close);
 
    sensor_msgs::PointCloud2 cloud_msg;
    pcl::toROSMsg(*remove_close, cloud_msg);
    in_publisher.publish(cloud_msg);
}
  这个框架： １．头文件.h class private:订阅发布话题，pointcb，其他点云处理函数的定义；public:PclTestCore(ros::NodeHandle &nh),~PclTestCore(),void spin()   ２．core PclTestCore::PclTestCore(ros::NodeHandel &nh) {订阅发布话题具体的，ros::spin()} PclTestCore::~PclTestCore(){} void PclTestCore::spin(){} void PclTestCore::其他函数具体定义 void PclTestCore::point_cb()   3._nodeint main(int argc,char **argv) {节点初始化，实例化，core，return}   ４）建立功能包，编译 功能包ninety(download baiduyun).编译 解决了找不到头文件的问题 写个launch文件 ninety.launch(mingtianxie) 或者直接   
roscore
rosbag play --clock test.bag
rosrun ninety ninety_node
rosrun rviz rviz
fixed_frame velodyne
add pointcloud2 /cloud_msg
  在这里插入图片描述