自学教程：C++ tree函数代码示例

int PointCloudStitching::stitch(pcl::PointCloud<PointT>& points, double epsilon, double maxCorrespondanceDistance){	if (this->stitching.size() == 0)	{		pcl::copyPointCloud(points, this->stitching);		return 0;	}	//	Compute surface normals and curvature	pcl::PointCloud<PointT> tempTarget;	pcl::copyPointCloud(this->stitching, tempTarget);	pcl::PointCloud<pcl::PointNormal>::Ptr pointsWithNormalSource (new pcl::PointCloud<pcl::PointNormal>);	pcl::PointCloud<pcl::PointNormal>::Ptr pointsWithNormalTarget (new pcl::PointCloud<pcl::PointNormal>);	pcl::NormalEstimation<PointT, pcl::PointNormal> normalEstimate;	pcl::search::KdTree<PointT>::Ptr tree (new pcl::search::KdTree<PointT> ());	normalEstimate.setSearchMethod(tree);	normalEstimate.setKSearch(30);	normalEstimate.setInputCloud(points.makeShared());	normalEstimate.compute(*pointsWithNormalSource);	pcl::copyPointCloud(points, *pointsWithNormalSource);	normalEstimate.setInputCloud (tempTarget.makeShared());	normalEstimate.compute(*pointsWithNormalTarget);	pcl::copyPointCloud (tempTarget, *pointsWithNormalTarget);	//	Instantiate custom point representation	MyPointNormal pointNormal;	//	... and weight the 'curvature' dimension so that it is balanced against x, y, and z	float alpha[4] = {1.0, 1.0, 1.0, 1.0};	pointNormal.setRescaleValues(alpha);	//	Align	pcl::IterativeClosestPointNonLinear<pcl::PointNormal, pcl::PointNormal> registration;	registration.setTransformationEpsilon(epsilon);	//	Set the maximum distance between two correspondences	registration.setMaxCorrespondenceDistance(maxCorrespondanceDistance);	//	Set the point representation	registration.setPointRepresentation (boost::make_shared<const MyPointNormal> (pointNormal));	registration.setInputSource(pointsWithNormalSource);	registration.setInputTarget(pointsWithNormalTarget);	registration.setMaximumIterations(30);	PCL_ERROR("Source size: %d  --  Target size: %d/n", (int)pointsWithNormalSource.get()->size(), (int)pointsWithNormalTarget.get()->size());	Eigen::Matrix4f tf = Eigen::Matrix4f::Identity();	pcl::PointCloud<pcl::PointNormal>::Ptr regResult = pointsWithNormalSource;	PCL_ERROR("Stitching ... ");	registration.align(*regResult);	PCL_ERROR("Done!/n");	tf = registration.getFinalTransformation().inverse();//	PCL_ERROR("/nTransform:/n");//	PCL_ERROR("| %f %f %f %f |/n", tf(0,0), tf(0,1), tf (0,2), tf(0,3));//	PCL_ERROR("| %f %f %f %f |/n", tf(1,0), tf(1,1), tf (1,2), tf(1,3));//	PCL_ERROR("| %f %f %f %f |/n", tf(2,0), tf(2,1), tf (2,2), tf(2,3));//	PCL_ERROR("| %f %f %f %f |/n/n", tf(3,0), tf(3,1), tf (3,2), tf(3,3));	pcl::transformPointCloud(*pointsWithNormalSource, *regResult, tf);	*regResult += *pointsWithNormalTarget;	pcl::copyPointCloud(*regResult, this->stitching);	return 0;}

  unsigned int scale_of_anisotropy (const Point_set& points, double& size)  {    Tree tree(points.begin(), points.end());        double ratio_kept = (points.size() < 1000)      ? 1. : 1000. / (points.size());        std::vector<Point> subset;    for (std::size_t i = 0; i < points.size (); ++ i)      if (rand() / (double)RAND_MAX < ratio_kept)    	subset.push_back (points[i]);        std::vector<unsigned int> scales;    generate_scales (std::back_inserter (scales));    std::vector<unsigned int> chosen;    for (std::size_t i = 0; i < subset.size (); ++ i)      {    	Neighbor_search search(tree, subset[i],scales.back());	double current = 0.;    	unsigned int nb = 0;    	std::size_t index = 0;	double maximum = 0.;	unsigned int c = 0;	    	for (Search_iterator search_iterator = search.begin();    	     search_iterator != search.end (); ++ search_iterator, ++ nb)    	  {	    current += search_iterator->second;    	    if (nb + 1 == scales[index])    	      {		double score = std::sqrt (current / scales[index])		  / std::pow (scales[index], 0.75); // NB ^ (3/4)		if (score > maximum)		  {		    maximum = score;		    c = scales[index];		  }    		++ index;    		if (index == scales.size ())    		  break;    	      }    	  }	chosen.push_back (c);      }    double mean = 0.;    for (std::size_t i = 0; i < chosen.size(); ++ i)      mean += chosen[i];    mean /= chosen.size();    unsigned int aniso_scale = static_cast<unsigned int>(mean);    size = 0.;    for (std::size_t i = 0; i < subset.size (); ++ i)      {    	Neighbor_search search(tree, subset[i], aniso_scale);	size += std::sqrt ((-- search.end())->second);      }    size /= subset.size();        return aniso_scale;  }

//***********************************************************// segmentation//  セグメンテ
C++ treeScope函数代码示例
C++ trayIconActivated函数代码示例