PCL（常用api）

这个博客持续更新啦---------------------------------

另外多参考这个博主的内容：https://www.cnblogs.com/flyinggod/p/9478000.html（写的很好啦）

1.找到目标点云中x,y,z中得最值，以及对应得索引；

		subset1_maxz = cloud_subset1->points[0].z;
		subset2_maxz = cloud_subset2->points[0].z;
		index_maxz1 = 0;
		index_maxz2 = 0;			

for (int l = 0; l < cloud_subset1->points.size(); l++)
		{
			if (subset1_maxz< cloud_subset1->points[l].z)
			{
				subset1_maxz = cloud_subset1->points[l].z;
				index_maxz1 = l;
			}
		}
		cloud_topboundary->push_back(cloud_subset1->points[index_maxz1]);
		

2.还有一种调用api的，但是好像没有返回索引的

	pcl::getMinMax3D(*cloud_final, min, max);
	cout << min.y << endl;
	cout << max.y << endl;

3.获得已知某一点（你想要的，例如点云平面的中心点）的索引的方法

（注意这里不初始化点云，直接赋值会报错）

		   pcl::PointXY dian;	
		   dian.x = (min.x+max.x) / 2;
		   dian.y = (min.y+max.y) / 2;

		   int index;
		   /*报错 原因估计是没有初始化 kk 结构*/
		   pcl::PointCloud<pcl::PointXY>::Ptr kk(new pcl::PointCloud<pcl::PointXY>);
		   kk->resize(cloud_cluster->points.size());
		   for (size_t i = 0; i < cloud_cluster->points.size(); i++)
		   {
			  
			   kk->points[i].x= cloud_cluster->points[i].x;
			   kk->points[i].y = cloud_cluster->points[i].y;

		   }
		   //找到索引啦
		   for (size_t i = 0; i < cloud_cluster->points.size(); i++)
		   {
			   float juli;
			   juli= pcl::squaredEuclideanDistance(dian, kk->points[0]);
			   if (juli < pcl::squaredEuclideanDistance(dian, kk->points[i]))
			   {
				   juli = pcl::squaredEuclideanDistance(dian, kk->points[i]);
				   index = i;

			   }

		   }

4.获得点云normal 值的方法（三种不同的方式）

pcl::PointCloud<pcl::Normal>::Ptr normals (new pcl::PointCloud<pcl::Normal>); 
std::cout<<normals->points[100]<<std::endl;
std::cout<<"["<<normals->points[100].normal_x<<" "
              <<normals->points[100].normal_y<<" "
              <<normals->points[100].normal_z<<" "
              <<normals->points[100].curvature<<"]"<<endl;

std::cout<<"["<<normals->points[100].normal[0]<<" "
              <<normals->points[100].normal[1]<<" "
              <<normals->points[100].normal[2]<<" "
              <<normals->points[100].curvature<<"]"<<endl;

std::cout<<"["<<normals->points[100].data_n[0]<<" "
              <<normals->points[100].data_n[1]<<" "
              <<normals->points[100].data_n[2]<<" "
              <<normals->points[100].curvature<<"]"<<endl;

 

5.时间计算（主要是要计算每一部分计算的时间） 

time_t start, end, diff[5], option;															 //时间变量
	start = time(0);
	printf("开始计时\n");




diff[0] = difftime(end, start);
	PCL_INFO("\ the time of load pointcloud: %d\n", diff[0]);



	diff[1] = difftime(end, start) - diff[0];
	PCL_INFO("\computing Normals takes(seconds): %d\n", diff[1]);