《PCL》获取点云边界

两个限制条件：

(M_PI_2);    //角度：
(atoi(argv[3]));  //搜索点：

#include <iostream>
#include <vector>
#include <ctime>
#include <boost/thread/>
#include <pcl/io/pcd_io.h>
#include <pcl/visualization/pcl_visualizer.h>
#include <pcl/console/>
#include <pcl/features/>
#include <pcl/features/>
#include <pcl/features/normal_3d.h>
#include <pcl/impl/point_types.hpp>
#include <pcl/features/>
#include <pcl/visualization/cloud_viewer.h>
using namespace std;

int main(int argc, char **argv)
{
	pcl::PointCloud<pcl::PointXYZ>::Ptr cloud(new pcl::PointCloud<pcl::PointXYZ>);
	
	if (pcl::io::loadPCDFile<pcl::PointXYZ>(argv[1], *cloud) == -1)
	{
		PCL_ERROR("COULD NOT READ FILE  \n");
		return (-1);
	}

	std::cout << "points sieze is:" << cloud->size() << std::endl;
	pcl::PointCloud<pcl::Normal>::Ptr normals(new pcl::PointCloud<pcl::Normal>);
	pcl::PointCloud<pcl::Boundary> boundaries;
	pcl::BoundaryEstimation<pcl::PointXYZ, pcl::Normal, pcl::Boundary> est;
	pcl::search::KdTree<pcl::PointXYZ>::Ptr tree(new pcl::search::KdTree<pcl::PointXYZ>());
	
	//创建一个快速k近邻查询,查询的时候若该点在点云中，则第一个近邻点是其本身
	pcl::KdTreeFLANN<pcl::PointXYZ> kdtree;  
	(cloud);
	int k =2;
	float everagedistance =0;
	for (int i =0; i < cloud->size()/2;i++)
	{
		//std::cout << "cloud->size()/2" << cloud->points[i] << std::endl;
		vector<int> nnh;
		vector<float> squaredistance;
		//  pcl::PointXYZ p;
		//   p = cloud->points[i];
		(cloud->points[i], k, nnh, squaredistance);
		/*std::cout << "查询点位： " << cloud->points[i] << std::endl;
		std::cout << "近邻为： " << nnh[0] << "  " << nnh[1] << std::endl;
		std::cout << "近邻为： " << cloud->points[nnh[0]] << "  " << cloud->points[nnh[1]] << std::endl;
*/
		everagedistance += sqrt(squaredistance[1]);
		//   cout<<everagedistance<<endl;
	
	}
	everagedistance = everagedistance/(cloud->size()/2);
	cout<<"everage distance is : "<<everagedistance<<endl;

	pcl::NormalEstimation<pcl::PointXYZ, pcl::Normal> normEst;  //其中pcl::PointXYZ表示输入类型数据，pcl::Normal表示输出类型,且pcl::Normal前三项是法向，最后一项是曲率
	(cloud);
	(tree);
	// (2);  //法向估计的半径
	(9);  //法向估计的点数
	(*normals);
	cout << "normal size is " << normals->size() << endl;

	//normal_est.setViewPoint(0,0,0); //这个应该会使法向一致
	(cloud);
	(normals);/*M_PI_2 */
	(M_PI_2);   ///在这里 由于构造函数已经对其进行了初始化 为Π/2 ，必须这样 使用 M_PI/2  M_PI_2  
	(tree);
	(atoi(argv[3]));  //一般这里的数值越高，最终边界识别的精度越好 20
	//  (everagedistance);  //搜索半径
	(boundaries);

	//  pcl::PointCloud<pcl::PointXYZ> boundPoints;
	pcl::PointCloud<pcl::PointXYZ>::Ptr boundPoints(new pcl::PointCloud<pcl::PointXYZ>);
	pcl::PointCloud<pcl::PointXYZ> noBoundPoints;
	int countBoundaries = 0;
	for (int i = 0; i<cloud->size(); i++){
		uint8_t x = ([i].boundary_point);
		int a = static_cast<int>(x); //该函数的功能是强制类型转换
		if (a == 1)
		{
			//  boundPoints.push_back(cloud->points[i]);
			(*boundPoints).push_back(cloud->points[i]);
			countBoundaries++;
		}
		else
			noBoundPoints.push_back(cloud->points[i]);

	}
	std::cout << "boudary size is：" << countBoundaries << std::endl;
	//  pcl::io::savePCDFileASCII("",boundPoints);

	pcl::io::savePCDFileASCII("", *boundPoints);
	pcl::io::savePCDFileASCII("", noBoundPoints);
	pcl::visualization::CloudViewer viewer("test");
	(boundPoints);
	while (!())
	{
	}
	return 0;
}