[C++知识库]点云处理常用指令记录（c++、pcl）

1.获得文件名的后缀

	std::string file_extension = model_filename_.substr(model_filename_.find_last_of('.'));

2.STL文件转pointcloud 可用github上的fast-ppf里面的meshsampling
3.获得点云的最大x，y，z及平均值

	pcl::PointXYZ minPt, maxPt, avgPt;
	pcl::getMinMax3D(*model_sampling, minPt, maxPt);
	avgPt.x = (minPt.x + maxPt.x) / 2;
	avgPt.y = (minPt.y + maxPt.y) / 2;
	avgPt.z = (minPt.z + maxPt.z) / 2;

4.从点云的正上中下共六个视角获得点云的放射点可以用可参考github上的fast-ppf里面的HPR。
5.点云显示更新

    CustomVisualizer customViewer
    customViewer.viewer->removeAllShapes();
	customViewer.viewer->removeAllPointClouds();
	customViewer.viewer->addPointCloud(model);
	std::getchar();
//先删内容，在增加内容，设置一个键盘输入

6.计算点云pointcloud的中心点值，并进行点云中心化变换

Eigen::Vector3d sum_of_pos = Eigen::Vector3d::Zero();
for (const auto& p : *(model)) sum_of_pos += p.getVector3fMap().cast<double>();

Eigen::Matrix4d transform_centering = Eigen::Matrix4d::Identity();
transform_centering.topRightCorner<3, 1>() = -sum_of_pos / model->size();
pcl::transformPointCloud(*model, *model, transform_centering);
pcl::transformPointCloud(*model, *model, Eigen::Vector3f(0, 0, 0), Eigen::Quaternionf(0.7071, 0, -0.7071, 0));

7.计算点云分辨率

double computeCloudResolution (const pcl::PointCloud<PointType>::ConstPtr &cloud)
  {
    double res = 0.0;
    int n_points = 0;
    int nres;
    std::vector<int> indices (2);
    std::vector<float> sqr_distances (2);
    pcl::search::KdTree<PointType> tree;
    tree.setInputCloud (cloud);
 
   for (size_t i = 0; i < cloud->size (); ++i)
   {
     if (! pcl_isfinite ((*cloud)[i].x))
     {
       continue;
     }
     //Considering the second neighbor since the first is the point itself.
     nres = tree.nearestKSearch (i, 2, indices, sqr_distances);
     if (nres == 2)
     {
       res += sqrt (sqr_distances[1]);
       ++n_points;
     }
   }
   if (n_points != 0)
   {
     res /= n_points;
   }
   return res;
3 }

8.计算点云最大直径

double computeCloudDiameter(const pcl::PointCloud<pcl::PointXYZ>::ConstPtr &cloud)
{
	pcl::PointCloud<pcl::PointXYZ>::Ptr cloud_downsampled(new pcl::PointCloud<pcl::PointXYZ>());
	pcl::VoxelGrid<PointXYZ> vg;
	vg.setInputCloud(cloud);
	vg.setLeafSize(0.005f, 0.005f, 0.005f);
	vg.setDownsampleAllData(false);
	vg.filter(*cloud_downsampled);

	double diameter_sqr = 0;
	for (size_t i = 0; i < cloud_downsampled->points.size(); i += 10)
	{
		for (size_t j = 0; j < cloud_downsampled->points.size(); j += 10)
		{
			if (i == j)
				continue;
			double distance_sqr = (cloud_downsampled->points[i].x - cloud_downsampled->points[j].x)*(cloud_downsampled->points[i].x - cloud_downsampled->points[j].x)
				+ (cloud_downsampled->points[i].y - cloud_downsampled->points[j].y)*(cloud_downsampled->points[i].y - cloud_downsampled->points[j].y)
				+ (cloud_downsampled->points[i].z - cloud_downsampled->points[j].z)*(cloud_downsampled->points[i].z - cloud_downsampled->points[j].z);
			if (distance_sqr > diameter_sqr)
			{
				diameter_sqr = distance_sqr;
			}
		}
	}
	return sqrt(diameter_sqr);
}

9.计算法向量

pcl::PointCloud<pcl::Normal>::Ptr normals(new pcl::PointCloud<pcl::Normal>);
pcl::NormalEstimationOMP<pcl::PointXYZ, pcl::Normal> ne;
pcl::search::KdTree<pcl::PointXYZ>::Ptr tree(new pcl::search::KdTree<pcl::PointXYZ>);
ne.setInputCloud(model_keypoints);
ne.setSearchSurface(model);
ne.setNumberOfThreads(8);
ne.setSearchMethod(tree);
ne.setRadiusSearch(norm_rad);
ne.compute(*normals);

10.
11.pcl::concatenateFields(*model_keypoints, *normals, *model_keypoints_with_normals);把法向量加到点云数据里面，生产带有法向量的点云，即pcl::PointNormal类型， 背景有XYZ，还有normal