1.计算FPFH描述子
#include <pcl/point_types.h>
#include <pcl/features/fpfh.h>
#include<pcl/io/pcd_io.h>
#include<iostream>
#include <pcl/features/normal_3d.h>
#include <pcl/visualization/pcl_plotter.h>// 直方图的可视化
using namespace std;
int main()
{
pcl::PointCloud<pcl::PointXYZ>::Ptr cloud(new pcl::PointCloud<pcl::PointXYZ>);
if (pcl::io::loadPCDFile("source.pcd", *cloud) == -1)
{
cout << "加载失败" << endl;
return -1;
}
//pcl::PointCloud<pcl::Normal>::Ptr normals(new pcl::PointCloud<pcl::Normal>());
//计算法向量
pcl::NormalEstimation<pcl::PointXYZ, pcl::Normal> ne;
ne.setInputCloud(cloud);
//创建一个空的kdtree对象，并把它传递给法线估计对象
//基于给出的输入数据集，kdtree将被建立
pcl::search::KdTree<pcl::PointXYZ>::Ptr tree(new pcl::search::KdTree<pcl::PointXYZ>());
ne.setSearchMethod(tree);
//输出数据集
pcl::PointCloud<pcl::Normal>::Ptr cloud_normals(new pcl::PointCloud<pcl::Normal>);
//使用半径在查询点周围3厘米范围内的所有邻元素
ne.setRadiusSearch(0.2);
//计算特征值
ne.compute(*cloud_normals);
// Create the FPFH estimation class, and pass the input dataset+normals to it
pcl::FPFHEstimation<pcl::PointXYZ, pcl::Normal, pcl::FPFHSignature33> fpfh;
fpfh.setInputCloud(cloud);
fpfh.setInputNormals(cloud_normals);
// alternatively, if cloud is of tpe PointNormal, do fpfh.setInputNormals (cloud);
// Create an empty kdtree representation, and pass it to the FPFH estimation object.
// Its content will be filled inside the object, based on the given input dataset (as no other search surface is given).
//pcl::search::KdTree<pcl::PointXYZ>::Ptr tree(new pcl::search::KdTree<pcl::PointXYZ>);
fpfh.setSearchMethod(tree);
// Output datasets
pcl::PointCloud<pcl::FPFHSignature33>::Ptr fpfhs(new pcl::PointCloud<pcl::FPFHSignature33>());
// Use all neighbors in a sphere of radius 5cm
// IMPORTANT: the radius used here has to be larger than the radius used to estimate the surface normals!!!
fpfh.setRadiusSearch(0.2);
// Compute the features
fpfh.compute(*fpfhs);
cout << "计算完毕";
}2.fpfh对应关系
#include <pcl/io/pcd_io.h>
#include <pcl/point_types.h>
#include <pcl/point_cloud.h>
#include <pcl/features/normal_3d_omp.h>
#include <pcl/features/fpfh_omp.h> // fpfh加速计算的omp(多核并行计算)
#include <pcl/registration/correspondence_estimation.h> // 获取对应关系的基类
#include <pcl/visualization/pcl_visualizer.h>
#include <boost/thread/thread.hpp>
#include<pcl/filters/filter.h>
#include <pcl/filters/voxel_grid.h>
pcl::PointCloud<pcl::PointXYZ> voxelGrid(pcl::PointCloud<pcl::PointXYZ>::Ptr cloud_in, float leaf)
{
pcl::VoxelGrid<pcl::PointXYZ> sor;
sor.setInputCloud(cloud_in);
sor.setLeafSize(leaf, leaf, leaf);
pcl::PointCloud<pcl::PointXYZ> cloud_out;
sor.filter(cloud_out);
return cloud_out;
}
using namespace std;
typedef pcl::PointCloud<pcl::FPFHSignature33> fpfhFeature;
fpfhFeature::Ptr compute_fpfh_feature(pcl::PointCloud<pcl::PointXYZ>::Ptr input_cloud)
{
pcl::search::KdTree<pcl::PointXYZ>::Ptr tree(new pcl::search::KdTree<pcl::PointXYZ>());
//-------------------------法向量估计-----------------------
pcl::PointCloud<pcl::Normal>::Ptr normals(new pcl::PointCloud<pcl::Normal>);
pcl::NormalEstimationOMP<pcl::PointXYZ, pcl::Normal> n;
n.setInputCloud(input_cloud);
n.setNumberOfThreads(8);//设置openMP的线程数
n.setSearchMethod(tree);
n.setKSearch(10);
n.compute(*normals);
//------------------FPFH估计-------------------------------
fpfhFeature::Ptr fpfh(new fpfhFeature);
pcl::FPFHEstimationOMP<pcl::PointXYZ, pcl::Normal, pcl::FPFHSignature33> f;
f.setNumberOfThreads(8); // 指定8核计算
f.setInputCloud(input_cloud);
f.setInputNormals(normals);
f.setSearchMethod(tree);
f.setKSearch(10);
f.compute(*fpfh);
return fpfh;
}
int
main(int argc, char** argv)
{
pcl::PointCloud<pcl::PointXYZ>::Ptr source_cloud(new pcl::PointCloud<pcl::PointXYZ>);
pcl::PointCloud<pcl::PointXYZ>::Ptr target_cloud(new pcl::PointCloud<pcl::PointXYZ>);
pcl::io::loadPCDFile<pcl::PointXYZ>("source.pcd", *source_cloud);
std::cout << "source:" << source_cloud->size() << endl;
pcl::io::loadPCDFile<pcl::PointXYZ>("target.pcd", *target_cloud);
std::cout << "target:" << target_cloud->size() << endl;
std::vector<int>nan_indicies;
std::vector<int>nan_indicies2;
pcl::removeNaNFromPointCloud(*source_cloud, *source_cloud, nan_indicies);
pcl::removeNaNFromPointCloud(*target_cloud, *target_cloud, nan_indicies2);
pcl::PointCloud<pcl::PointXYZ>::Ptr vox_source_cloud=voxelGrid(source_cloud, 0.7).makeShared();//下采样后的点云
pcl::PointCloud<pcl::PointXYZ>::Ptr vox_target_cloud = voxelGrid(target_cloud, 0.7).makeShared();
std::cout << "source_voxed"<<vox_source_cloud->size() << endl;
std::cout << "target_voxed" << vox_target_cloud->size() << endl;
//---------------------计算源点云和目标点云的FPFH-----------------
fpfhFeature::Ptr source_fpfh = compute_fpfh_feature(vox_source_cloud);
fpfhFeature::Ptr target_fpfh = compute_fpfh_feature(vox_target_cloud);
//--------------------根据特征描述子获取对应关系------------------
pcl::registration::CorrespondenceEstimation<pcl::FPFHSignature33, pcl::FPFHSignature33> crude_cor_est;
boost::shared_ptr<pcl::Correspondences> cru_correspondences(new pcl::Correspondences);
crude_cor_est.setInputSource(source_fpfh);
crude_cor_est.setInputTarget(target_fpfh);
// crude_cor_est.determineCorrespondences(cru_correspondences,1); // 获取对应关系，第二个参数为对应关系之间的最大距离阈值
crude_cor_est.determineReciprocalCorrespondences(*cru_correspondences);// 相互查找的方式获取对应关系
std::cout << "source:" << source_cloud->size() << endl;
cout << "输入源点云：" << source_cloud->size() << "个" << endl;
cout << "输入源点云：" << source_cloud->points.size() << "个" << endl;
cout << "输入目标点云点云：" << target_cloud->points.size() << "个" << endl;
cout << "获取匹配点对:" << cru_correspondences->size() << endl;
//---------------------可视化点云及对应关系----------------------
boost::shared_ptr<pcl::visualization::PCLVisualizer>viewer(new pcl::visualization::PCLVisualizer("显示点云"));
viewer->setBackgroundColor(0, 0, 0); //设置背景颜色为黑色
// 对目标点云着色可视化 (red).
pcl::visualization::PointCloudColorHandlerCustom<pcl::PointXYZ>target_color(target_cloud, 255, 0, 0);
viewer->addPointCloud<pcl::PointXYZ>(target_cloud, target_color, "target");
viewer->setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 1, "target");
// 对源点云着色可视化 (green).
pcl::visualization::PointCloudColorHandlerCustom<pcl::PointXYZ>source_color(source_cloud, 0, 255, 0);
viewer->addPointCloud<pcl::PointXYZ>(source_cloud, source_color, "source");
viewer->setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 1, "source");
//对应关系可视化
viewer->setWindowName("基于特征描述子的对应");
viewer->addCorrespondences<pcl::PointXYZ>(source_cloud, target_cloud, *cru_correspondences, "correspondence");
//viewer->initCameraParameters();
while (!viewer->wasStopped())
{
viewer->spinOnce(100);
boost::this_thread::sleep(boost::posix_time::microseconds(100000));
}
return 0;
}//fpfh显示对应关系原点云个数为什么会变，不太清楚
（1）点对对应关系函数
（pcl::registration::CorrespondenceEstimation<pcl::PointXYZ, pcl::PointXYZ>core;）
#include<pcl/io/pcd_io.h>
#include<pcl/kdtree/io.h>
#include <pcl/filters/extract_indices.h>
#include<pcl/registration/correspondence_estimation.h>
using namespace std;
int main()
{
pcl::PointCloud<pcl::PointXYZ>::Ptr source(new pcl::PointCloud<pcl::PointXYZ>);
pcl::PointCloud<pcl::PointXYZ>::Ptr target(new pcl::PointCloud<pcl::PointXYZ>);
//----------------------读取点云-------------------------
pcl::io::loadPCDFile<pcl::PointXYZ>("source.pcd", *source);
pcl::io::loadPCDFile<pcl::PointXYZ>("target.pcd", *target);
//--------------------提取重叠部分-----------------------
pcl::registration::CorrespondenceEstimation<pcl::PointXYZ, pcl::PointXYZ>core;
core.setInputSource(source);
core.setInputTarget(target);
boost::shared_ptr<pcl::Correspondences> cor(new pcl::Correspondences);
core.determineCorrespondences(*cor, 0.5); //对应点之间的最大距离
double mean = 0.0, stddev = 0.0;
pcl::registration::getCorDistMeanStd(*cor, mean, stddev);
cout << "对应点距离的均值为：" << mean << "\n"
<< "对应点距离平方标准差为：" << stddev << endl;
return 0;
}//对应点对距离平方的均值和方差（2）基于距离获取对应关系
#include <iostream>
#include <pcl/io/pcd_io.h>
#include <pcl/point_types.h>
#include <boost/thread/thread.hpp>
#include <pcl/visualization/pcl_visualizer.h>
#include <pcl/registration/correspondence_estimation.h> // 获取对应关系的基类
#include <pcl/registration/correspondence_rejection_distance.h> // 根据距离筛选对应点对
using namespace std;
int
main(int argc, char** argv)
{
// 加载目标点云
pcl::PointCloud<pcl::PointXYZ>::Ptr target(new pcl::PointCloud<pcl::PointXYZ>);
pcl::io::loadPCDFile<pcl::PointXYZ>("plant.pcd", *target);
cout << "从目标点云中读取 " << target->size() << " 个点" << endl;
// 加载源点云
pcl::PointCloud<pcl::PointXYZ>::Ptr source(new pcl::PointCloud<pcl::PointXYZ>);
pcl::io::loadPCDFile<pcl::PointXYZ>("plant1.pcd", *source);
cout << "从源点云中读取 " << source->size() << " 个点" << endl;
//-------------------获取对应关系---------------------------
pcl::CorrespondencesPtr correspondences(new pcl::Correspondences);
pcl::registration::CorrespondenceEstimation<pcl::PointXYZ, pcl::PointXYZ> corr_est;
corr_est.setInputSource(source);
corr_est.setInputTarget(target);
corr_est.determineCorrespondences(*correspondences); // 获取对应关系
//----------------基于距离筛选对应关系----------------------
pcl::CorrespondencesPtr correspondences_result_rej_dist(new pcl::Correspondences);
pcl::registration::CorrespondenceRejectorDistance corr_rej_dist;
corr_rej_dist.setInputCorrespondences(correspondences);// 输入对应关系
corr_rej_dist.setMaximumDistance(0.1); // 对应关系之间的最大距离阈值
corr_rej_dist.getCorrespondences(*correspondences_result_rej_dist); // 输出筛选后的对应关系
cout << "获取对应点对：" << correspondences->size() << "个" << endl;
cout << "基于距离筛选后还剩下：" << correspondences_result_rej_dist->size() << "个" << endl;
//-----------------可视化点云及对应关系--------------------
boost::shared_ptr<pcl::visualization::PCLVisualizer>viewer(new pcl::visualization::PCLVisualizer("显示点云"));
viewer->setBackgroundColor(0, 0, 0); // 设置背景颜色为黑色
// 对目标点云着色可视化 (red).
pcl::visualization::PointCloudColorHandlerCustom<pcl::PointXYZ>target_color(target, 255, 0, 0);
viewer->addPointCloud<pcl::PointXYZ>(target, target_color, "target");
viewer->setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 1, "target");
// 对源点云着色可视化 (green).
pcl::visualization::PointCloudColorHandlerCustom<pcl::PointXYZ>input_color(source, 0, 255, 0);
viewer->addPointCloud<pcl::PointXYZ>(source, input_color, "source");
viewer->setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 1, "source");
//对应关系可视化
viewer->setWindowName("基于距离的对应拒绝");
viewer->addCorrespondences<pcl::PointXYZ>(source, target, *correspondences_result_rej_dist, "correspondence");
//viewer->initCameraParameters();
while (!viewer->wasStopped())
{
viewer->spinOnce(100);
boost::this_thread::sleep(boost::posix_time::microseconds(100000));
}
system("pause");
return 0;
}//基于距离获取对应关系（3）获取对应关系
#include <iostream>
#include <pcl/point_types.h>
#include <pcl/io/pcd_io.h>
#include <pcl/features/normal_3d_omp.h>
#include <pcl/registration/correspondence_estimation_normal_shooting.h>
#include <pcl/registration/correspondence_rejection_surface_normal.h>
#include <pcl/registration/transformation_estimation_lm.h>
#include <boost/thread/thread.hpp>
#include <pcl/visualization/pcl_visualizer.h>
using namespace std;
void compute_normal(pcl::PointCloud<pcl::PointXYZ>::Ptr& cloud, pcl::PointCloud<pcl::PointNormal>::Ptr& normals)
{
pcl::NormalEstimationOMP<pcl::PointXYZ, pcl::PointNormal> n;//OMP加速
//建立kdtree来进行近邻点集搜索
pcl::search::KdTree<pcl::PointXYZ>::Ptr tree(new pcl::search::KdTree<pcl::PointXYZ>());
n.setNumberOfThreads(6);//设置openMP的线程数
//n.setViewPoint(0,0,0);//设置视点，默认为（0，0，0）
n.setInputCloud(cloud);
n.setSearchMethod(tree);
n.setKSearch(10);//点云法向计算时，需要所搜的近邻点大小
//n.setRadiusSearch(0.03);//半径搜素
n.compute(*normals);//开始进行法向计
}
int
main(int argc, char** argv)
{
// 加载源点云
pcl::PointCloud<pcl::PointXYZ>::Ptr source(new pcl::PointCloud<pcl::PointXYZ>);
if (pcl::io::loadPCDFile<pcl::PointXYZ>("plant.pcd", *source) == -1)
{
PCL_ERROR("读取目标点云失败 \n");
return (-1);
}
cout << "从目标点云中读取 " << source->size() << " 个点" << endl;
// ----------------------加载目标点云-----------------------
pcl::PointCloud<pcl::PointXYZ>::Ptr target(new pcl::PointCloud<pcl::PointXYZ>);
if (pcl::io::loadPCDFile<pcl::PointXYZ>("plant1.pcd", *target) == -1)
{
PCL_ERROR("读取源标点云失败 \n");
return (-1);
}
cout << "从源点云中读取 " << target->size() << " 个点" << endl;
//-------------------------法线估计--------------------------
pcl::PointCloud<pcl::PointNormal>::Ptr source_normals(new pcl::PointCloud<pcl::PointNormal>);
compute_normal(source, source_normals);
pcl::PointCloud<pcl::PointNormal>::Ptr target_normals(new pcl::PointCloud<pcl::PointNormal>);
compute_normal(target, target_normals);
//----------------------获取匹配点对-------------------------
pcl::registration::CorrespondenceEstimationNormalShooting<pcl::PointXYZ, pcl::PointXYZ, pcl::PointNormal>core;
core.setInputSource(source);
core.setSourceNormals(source_normals);
core.setInputTarget(target);
core.setKSearch(10);
pcl::CorrespondencesPtr all_correspondences(new pcl::Correspondences);
core.determineCorrespondences(*all_correspondences, 0.5);//输入的是源点云上法线与目标点云上对应点之间的最大距离
//core.determineReciprocalCorrespondences(*all_correspondences); //确定输入点云与目标点云之间的交互对应关系。
//-------------------剔除错误匹配点对-----------------------
pcl::registration::CorrespondenceRejectorSurfaceNormal rejector;
rejector.initializeDataContainer<pcl::PointNormal, pcl::PointNormal>();// 初始化点类型和普通类型的数据容器对象
rejector.setInputSource<pcl::PointNormal>(source_normals);
rejector.setInputTarget<pcl::PointNormal>(target_normals);
rejector.setInputNormals<pcl::PointNormal, pcl::PointNormal>(source_normals);
rejector.setTargetNormals<pcl::PointNormal, pcl::PointNormal>(target_normals);
rejector.setInputCorrespondences(all_correspondences);
float threshold = cos(M_PI * 10 / 180);//计算cos(10°)
rejector.setThreshold(threshold); // 设置法线之间的夹角阈值，输入的是夹角余弦值
pcl::CorrespondencesPtr correspondences_after_rejector(new pcl::Correspondences);
rejector.getCorrespondences(*correspondences_after_rejector);
//cout << "Correspondences (After) : " << correspondences_after_rejector->size() << "\n";
Eigen::Matrix4f transformation = Eigen::Matrix4f::Identity();
pcl::registration::TransformationEstimationLM<pcl::PointXYZ, pcl::PointXYZ> trans_est_lm;
trans_est_lm.estimateRigidTransformation(*source, *target, *correspondences_after_rejector, transformation);
pcl::PointCloud<pcl::PointXYZ>::Ptr out(new pcl::PointCloud<pcl::PointXYZ>);
pcl::transformPointCloud(*source, *out, transformation);
pcl::io::savePCDFile("left.pcd", *out);
cout << "原始匹配点对有：" << all_correspondences->size() << "对" << endl;
cout << "剔除错误点对后有：" << correspondences_after_rejector->size() << "对" << endl;
cout << "LM求解的变换矩阵为：\n" << transformation << endl;
//---------------------------可视化-------------------------
boost::shared_ptr<pcl::visualization::PCLVisualizer>viewer(new pcl::visualization::PCLVisualizer("显示点云"));
viewer->setBackgroundColor(0, 0, 0); //设置背景颜色为黑色
// 对源点云着色可视化 (green).
pcl::visualization::PointCloudColorHandlerCustom<pcl::PointXYZ>input_color(source, 0, 255, 0);
viewer->addPointCloud<pcl::PointXYZ>(source, input_color, "source");
viewer->setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 1, "source");
// 对目标点云着色可视化 (red).
pcl::visualization::PointCloudColorHandlerCustom<pcl::PointXYZ>target_color(target, 255, 0, 0);
viewer->addPointCloud<pcl::PointXYZ>(target, target_color, "target");
viewer->setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 1, "target");
//----------------------对应关系可视化---------------------
viewer->addCorrespondences<pcl::PointXYZ>(source, target, *correspondences_after_rejector, "correspondence");
viewer->initCameraParameters();
viewer->addText("RejectorSurfaceNormal", 10, 10, "text");
while (!viewer->wasStopped())
{
viewer->spinOnce(100);
boost::this_thread::sleep(boost::posix_time::microseconds(100000));
}
return 0;
}//对应关系（4）提取点云非重叠部分
#include<pcl/io/pcd_io.h>
#include<pcl/kdtree/io.h>
#include <pcl/filters/extract_indices.h>
#include<pcl/registration/correspondence_estimation.h>
#include<pcl/filters/filter.h>
using namespace std;
int main()
{
pcl::PointCloud<pcl::PointXYZ>::Ptr source(new pcl::PointCloud<pcl::PointXYZ>);
pcl::PointCloud<pcl::PointXYZ>::Ptr target(new pcl::PointCloud<pcl::PointXYZ>);
//----------------------读取点云-------------------------
pcl::io::loadPCDFile<pcl::PointXYZ>("source.pcd", *source);
pcl::io::loadPCDFile<pcl::PointXYZ>("target.pcd", *target);
//--------------------提取重叠部分-----------------------
std::vector<int>nan_indicies;
pcl::removeNaNFromPointCloud(*source, *source, nan_indicies);
pcl::removeNaNFromPointCloud(*target, *target, nan_indicies);
pcl::registration::CorrespondenceEstimation<pcl::PointXYZ, pcl::PointXYZ>core;
core.setInputSource(source);
core.setInputTarget(target);
boost::shared_ptr<pcl::Correspondences> cor(new pcl::Correspondences);
core.determineReciprocalCorrespondences(*cor, 0.1); //对应点之间的最大距离
//------------------构造重叠点云的索引------------------
vector<int>pointIdxVec_A;
vector<int>pointIdxVec_B;
pcl::registration::getQueryIndices(*cor, pointIdxVec_A); // 获取重叠部分对应于setInputSource()输入点云的索引
pcl::registration::getMatchIndices(*cor, pointIdxVec_B); // 获取重叠部分对应于setInputTarget()输入点云的索引
//------------------根据索引提取点云--------------------
pcl::PointCloud<pcl::PointXYZ>::Ptr non_overlapA(new pcl::PointCloud<pcl::PointXYZ>);
pcl::PointCloud<pcl::PointXYZ>::Ptr non_overlapB(new pcl::PointCloud<pcl::PointXYZ>);
int non_A = target->points.size() - pointIdxVec_B.size();
int non_B = source->points.size() - pointIdxVec_A.size();
if (non_A < non_B) {
pcl::ExtractIndices<pcl::PointXYZ> extrA;
extrA.setInputCloud(source);//设置输入点云
extrA.setIndices(std::make_shared<vector<int>>(pointIdxVec_A));//设置索引
extrA.setNegative(true); // 提取对应索引之外的点
extrA.filter(*non_overlapA);
pcl::io::savePCDFileASCII("non_overlap.pcd", *non_overlapA);
}
else {
pcl::ExtractIndices<pcl::PointXYZ> extrB;
extrB.setInputCloud(target);
extrB.setIndices(std::make_shared<vector<int>>(pointIdxVec_B));
extrB.setNegative(true);
extrB.filter(*non_overlapB);
pcl::io::savePCDFileASCII("non_overlap.pcd", *non_overlapB);
};
return 0;
}//点云非重叠部分3.Robust pose estimation of rigid objects
点云类型为PointNormal（pcl::SampleConsensusPrerejective<PointNT, PointNT, FeatureT> align;）
#include <Eigen/Core>
#include <pcl/point_types.h>
#include <pcl/point_cloud.h>
#include <pcl/common/time.h>
#include <pcl/console/print.h>
#include <pcl/features/normal_3d.h>
#include <pcl/features/fpfh.h>
#include <pcl/filters/filter.h>
#include <pcl/filters/voxel_grid.h>
#include <pcl/io/pcd_io.h>
#include <pcl/registration/icp.h>
#include <pcl/registration/sample_consensus_prerejective.h>
#include <pcl/segmentation/sac_segmentation.h>
#include <pcl/visualization/pcl_visualizer.h>
// Types
typedef pcl::PointNormal PointNT;
typedef pcl::PointCloud<PointNT> PointCloudT;
typedef pcl::FPFHSignature33 FeatureT;// FPFH特殊描述子
typedef pcl::FPFHEstimation<PointNT, PointNT, FeatureT> FeatureEstimationT;//FPFH特征估计类
typedef pcl::PointCloud<FeatureT> FeatureCloudT;
typedef pcl::visualization::PointCloudColorHandlerCustom<PointNT> ColorHandlerT;//自定义颜色句柄
// Align a rigid object to a scene with clutter and occlusions
int
main(int argc, char** argv)
{
// Point clouds
PointCloudT::Ptr object(new PointCloudT);//目标点云
PointCloudT::Ptr object_aligned(new PointCloudT);
PointCloudT::Ptr scene(new PointCloudT);//源点云
FeatureCloudT::Ptr object_features(new FeatureCloudT);//目标特征描述子
FeatureCloudT::Ptr scene_features(new FeatureCloudT);//源点云特征描述子
// Get input object and scene
if (argc != 3)
{
pcl::console::print_error("Syntax is: %s object.pcd scene.pcd\n", argv[0]);
return (1);
}
// 加载目标物体和场景点云
pcl::console::print_highlight("Loading point clouds...\n");
if (pcl::io::loadPCDFile<PointNT>(argv[1], *object) < 0 ||
pcl::io::loadPCDFile<PointNT>(argv[2], *scene) < 0)
{
std::vector<int>nan_indicidies;
pcl::removeNaNFromPointCloud(*object, *object,nan_indicidies);
pcl::removeNaNFromPointCloud(*scene, *scene,nan_indicidies);
pcl::console::print_error("Error loading object/scene file!\n");
return (1);
}
// 下采样
pcl::console::print_highlight("Downsampling...\n");
pcl::VoxelGrid<PointNT> grid;
const float leaf = 0.005f;
grid.setLeafSize(leaf, leaf, leaf);//设置滤波时创建的体素大小
grid.setInputCloud(object);//设置需要过滤的对象
grid.filter(*object);//执行滤波处理，存储
grid.setInputCloud(scene);//设置需要过滤的对象
grid.filter(*scene);//执行滤波处理，存储
// 估计场景法线
pcl::console::print_highlight("Estimating scene normals...\n");
pcl::NormalEstimation<PointNT, PointNT> nest;//法线估计初始化
nest.setRadiusSearch(0.01);//一般为点云分辨率的5倍以上
nest.setInputCloud(scene);
nest.compute(*scene);
// 特征估计
pcl::console::print_highlight("Estimating features...\n");
FeatureEstimationT fest;//特征估计
fest.setRadiusSearch(0.025);
fest.setInputCloud(object);
fest.setInputNormals(object);
fest.compute(*object_features);
fest.setInputCloud(scene);
fest.setInputNormals(scene);
fest.compute(*scene_features);
// 实施配准
pcl::console::print_highlight("Starting alignment...\n");
pcl::SampleConsensusPrerejective<PointNT, PointNT, FeatureT> align;
align.setInputSource(object);
align.setSourceFeatures(object_features);
align.setInputTarget(scene);
align.setTargetFeatures(scene_features);
align.setMaximumIterations(50000); // 采样一致性迭代次数
align.setNumberOfSamples(3); // 创建假设所需的样本数
align.setCorrespondenceRandomness(5); // 使用的临近特征点的数目
align.setSimilarityThreshold(0.9f); // 多边形边长度相似度阈值
align.setMaxCorrespondenceDistance(2.5f * 0.005); // 判断是否为内点的距离阈值
align.setInlierFraction(0.25f); //接受位姿假设所需的内点比例
{
pcl::ScopeTime t("Alignment");
align.align(*object_aligned);
}
if (align.hasConverged())
{
// Print results
printf("\n");
Eigen::Matrix4f transformation = align.getFinalTransformation();
pcl::console::print_info(" | %6.3f %6.3f %6.3f | \n", transformation(0, 0), transformation(0, 1), transformation(0, 2));
pcl::console::print_info("R = | %6.3f %6.3f %6.3f | \n", transformation(1, 0), transformation(1, 1), transformation(1, 2));
pcl::console::print_info(" | %6.3f %6.3f %6.3f | \n", transformation(2, 0), transformation(2, 1), transformation(2, 2));
pcl::console::print_info("\n");
pcl::console::print_info("t = < %0.3f, %0.3f, %0.3f >\n", transformation(0, 3), transformation(1, 3), transformation(2, 3));
pcl::console::print_info("\n");
pcl::console::print_info("Inliers: %i/%i\n", align.getInliers().size(), object->size());
// Show alignment
pcl::visualization::PCLVisualizer visu("点云库PCL学习教程第二版-鲁棒位姿估计");
int v1(0), v2(0);
visu.createViewPort(0, 0, 0.5, 1, v1);
visu.createViewPort(0.5, 0, 1, 1, v2);
visu.setBackgroundColor(255, 255, 255, v1);
visu.addPointCloud(scene, ColorHandlerT(scene, 0.0, 255.0, 0.0), "scene", v1);
visu.addPointCloud(object_aligned, ColorHandlerT(object_aligned, 0.0, 0.0, 255.0), "object_aligned", v1);
visu.addPointCloud(object, ColorHandlerT(object, 0.0, 255.0, 0.0), "object_before_aligned", v2);
visu.addPointCloud(scene, ColorHandlerT(scene, 0.0, 0.0, 255.0), "scene_v2", v2);
visu.setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 3, "scene");
visu.setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 3, "object_aligned");
visu.setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 3, "object_before_aligned");
visu.setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 3, "scene_v2");
visu.spin();
}
else
{
pcl::console::print_error("Alignment failed!\n");
return (1);
}
return (0);
}配准后使用icp
#include <Eigen/Core>
#include <pcl/point_types.h>
#include <pcl/point_cloud.h>
#include <pcl/io/pcd_io.h>
#include <pcl/common/time.h>//计算时间
#include <pcl/console/print.h>//PCL控制台输出
#include <pcl/features/normal_3d_omp.h>//使用OMP需要添加的头文件
#include <pcl/features/fpfh.h>
#include <pcl/filters/filter.h>
#include <pcl/filters/voxel_grid.h>
#include <pcl/registration/sample_consensus_prerejective.h>
#include <pcl/visualization/pcl_visualizer.h>
#include <pcl/registration/icp.h>
// Types
typedef pcl::PointXYZ PointT;
typedef pcl::PointCloud<PointT> PointCloud;
typedef pcl::PointNormal PointNT;
typedef pcl::PointCloud<PointNT> PointCloudT;
typedef pcl::FPFHSignature33 FeatureT;
typedef pcl::FPFHEstimation<PointNT, PointNT, FeatureT> FeatureEstimationT;
typedef pcl::PointCloud<FeatureT> FeatureCloudT;
typedef pcl::visualization::PointCloudColorHandlerCustom<PointNT> ColorHandlerT;
// 将一个刚性物体与一个带有杂波和遮挡的场景对齐
int
main(int argc, char** argv)
{
// Point clouds
PointCloudT::Ptr object(new PointCloudT);
PointCloudT::Ptr object_aligned(new PointCloudT);
PointCloudT::Ptr scene(new PointCloudT);
FeatureCloudT::Ptr object_features(new FeatureCloudT);
FeatureCloudT::Ptr scene_features(new FeatureCloudT);
// 加载目标物体和场景点云
pcl::console::print_highlight("加载点云数据\n");
if (pcl::io::loadPCDFile<PointNT>("target.pcd", *object) < 0 ||
pcl::io::loadPCDFile<PointNT>("source.pcd", *scene) < 0)
{
pcl::console::print_error("点云加载失败!\n");
return (1);
}
std::vector<int>nan_indicies;
pcl::removeNaNFromPointCloud(*object, *object, nan_indicies);
pcl::removeNaNFromPointCloud(*scene, *scene, nan_indicies);
std::cout << object->size()<<endl;
std::cout << scene->size() << endl;
// 下采样
pcl::console::print_highlight("点云下采样\n");
pcl::VoxelGrid<PointNT> grid;
const float leaf = 0.8f;
grid.setLeafSize(leaf, leaf, leaf);
grid.setInputCloud(object);
grid.filter(*object);
grid.setInputCloud(scene);
grid.filter(*scene);
std::cout << object->size() << endl;
std::cout << scene->size() << endl;
// 估计场景法线
pcl::console::print_highlight("估计场景点云的法线\n");
pcl::NormalEstimationOMP<PointNT, PointNT> n;
n.setNumberOfThreads(4);//设置openMP的线程数
n.setRadiusSearch(1.25);
n.setInputCloud(scene);
n.compute(*scene);
n.setInputCloud(object);
n.compute(*object);
// 特征估计
pcl::console::print_highlight("计算FPFH特征\n");
FeatureEstimationT f;
f.setRadiusSearch(1.25);
f.setInputCloud(object);
f.setInputNormals(object);
f.compute(*object_features);
f.setInputCloud(scene);
f.setInputNormals(scene);
f.compute(*scene_features);
// 实施配准
pcl::console::print_highlight("开始进行配准\n");
pcl::SampleConsensusPrerejective<PointNT, PointNT, FeatureT> align;
align.setInputSource(scene);
align.setSourceFeatures(scene_features);
align.setInputTarget(object);
align.setTargetFeatures(object_features);
align.setMaximumIterations(50000); // 采样一致性迭代次数
align.setNumberOfSamples(3); // 创建假设所需的样本数
align.setCorrespondenceRandomness(5); // 使用的临近特征点的数目
align.setSimilarityThreshold(0.8f); // 多边形边长度相似度阈值
align.setMaxCorrespondenceDistance(2.6f * 1.0); // 判断是否为内点的距离阈值
align.setInlierFraction(0.25f); // 接受位姿假设所需的内点比例
{
pcl::ScopeTime t("执行配准");
align.align(*object_aligned);
}
pcl::io::savePCDFileASCII("object_aligned.pcd", *object_aligned);
PointCloudT::Ptr icp_result(new PointCloudT);
pcl::IterativeClosestPoint<PointNT, PointNT> icp;
icp.setInputSource(object_aligned);
icp.setInputTarget(object);
icp.setMaxCorrespondenceDistance(1.5);
// 最大迭代次数
icp.setMaximumIterations(100);
icp.setTransformationEpsilon(1e-10);// 两次变化矩阵之间的差值
icp.setEuclideanFitnessEpsilon(0.02);// 均方误差前后两次迭代方差的插值
icp.align(*icp_result);
pcl::io::savePCDFile("icp_result.pcd", *icp_result);
std::cout << "all aligned!" << endl;
if (align.hasConverged())
{
// Print results
printf("\n");
Eigen::Matrix4f transformation = align.getFinalTransformation();
pcl::console::print_info(" | %6.3f %6.3f %6.3f | \n", transformation(0, 0), transformation(0, 1), transformation(0, 2));
pcl::console::print_info("R = | %6.3f %6.3f %6.3f | \n", transformation(1, 0), transformation(1, 1), transformation(1, 2));
pcl::console::print_info(" | %6.3f %6.3f %6.3f | \n", transformation(2, 0), transformation(2, 1), transformation(2, 2));
pcl::console::print_info("\n");
pcl::console::print_info("t = < %0.3f, %0.3f, %0.3f >\n", transformation(0, 3), transformation(1, 3), transformation(2, 3));
pcl::console::print_info("\n");
pcl::console::print_info("Inliers: %i/%i\n", align.getInliers().size(), object->size());
// Show alignment
pcl::console::print_highlight("左侧为配准前的位置，右侧为配准后\n");
pcl::visualization::PCLVisualizer visu("鲁棒位姿估计");
visu.initCameraParameters();
int v1(0), v2(0);
visu.setWindowName("鲁棒位姿估计");
visu.createViewPort(0, 0, 0.5, 1, v1);
visu.createViewPort(0.5, 0, 1, 1, v2);
visu.setBackgroundColor(255, 255, 255, v2);
visu.addPointCloud(scene, ColorHandlerT(scene, 0.0, 255.0, 0.0), "scene", v2);
visu.addPointCloud(object_aligned, ColorHandlerT(object_aligned, 0.0, 0.0, 255.0), "object_aligned", v2);
visu.addPointCloud(object, ColorHandlerT(object, 0.0, 255.0, 0.0), "object_before_aligned", v1);
visu.addPointCloud(scene, ColorHandlerT(scene, 0.0, 0.0, 255.0), "scene_v2", v1);
visu.setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 3, "scene");
visu.setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 3, "object_aligned");
visu.setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 3, "object_before_aligned");
visu.setPointCloudRenderingProperties(pcl::visualization::PCL_VISUALIZER_POINT_SIZE, 3, "scene_v2");
visu.spin();
}
else
{
pcl::console::print_error("配准失败!\n");
return (1);
}
return (0);
}RANSAC+FPFH点云输入类型为PointXYZ
#include <Eigen/Core>
#include <pcl/point_types.h>
#include <pcl/point_cloud.h>
#include <pcl/common/time.h>
#include <pcl/console/print.h>
#include <pcl/features/normal_3d_omp.h>
#include <pcl/features/fpfh_omp.h>
#include <pcl/filters/filter.h>
#include <pcl/filters/voxel_grid.h>
#include <pcl/io/io.h>
#include <pcl/io/pcd_io.h>
#include <pcl/registration/icp.h>
#include <pcl/registration/sample_consensus_prerejective.h>
#include <pcl/segmentation/sac_segmentation.h>
#include <pcl/visualization/pcl_visualizer.h>
#include<pcl/features/normal_3d.h>
#include <pcl/features/fpfh.h>
// Align a rigid object to a scene with clutter and occlusions
int
main(int argc, char** argv)
{
// Types
typedef pcl::PointNormal PointNT;
//typedef pcl::PointXYZ PointNT;
typedef pcl::PointCloud<PointNT> PointCloudT;
typedef pcl::FPFHSignature33 FeatureT;
typedef pcl::FPFHEstimationOMP<PointNT, PointNT, FeatureT> FeatureEstimationT;
typedef pcl::PointCloud<FeatureT> FeatureCloudT;
typedef pcl::visualization::PointCloudColorHandlerCustom<PointNT> ColorHandlerT;
// Point clouds
PointCloudT::Ptr object(new PointCloudT);
PointCloudT::Ptr object_aligned(new PointCloudT);
PointCloudT::Ptr scene(new PointCloudT);
PointCloudT::Ptr down_s(new PointCloudT);
PointCloudT::Ptr down_t(new PointCloudT);
PointCloudT::Ptr normal_s(new PointCloudT);
PointCloudT::Ptr normal_t(new PointCloudT);
FeatureCloudT::Ptr object_features(new FeatureCloudT);
FeatureCloudT::Ptr scene_features(new FeatureCloudT);
// Load object and scene
/*pcl::console::print_highlight("Loading point clouds...\n");
if (pcl::io::loadPCDFile<PointNT>("C:\\Users\\SCQ\\Desktop\\source.pcd", *object) < 0 ||
pcl::io::loadPCDFile<PointNT>("C:\\Users\\SCQ\\Desktop\\target.pcd", *scene) < 0)
{
pcl::console::print_error("Error loading object/scene file!\n");
return (1);
}*/
// Estimate normals for scene
pcl::console::print_highlight("Estimating scene normals...\n");
/*pcl::NormalEstimationOMP<PointNT, PointNT> nest;
nest.setRadiusSearch(0.2);
nest.setInputCloud(scene);
nest.compute(*normal_s);
nest.setRadiusSearch(0.5);
nest.setInputCloud(object);
nest.compute(*normal_t);
pcl::io::savePCDFile("normal22.pcd", *object);
pcl::io::savePCDFile("normal.pcd", *normal_t);*/
//!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!!
pcl::console::print_highlight("Loading point clouds...\n");
pcl::PointCloud<pcl::PointXYZ>::Ptr source(new pcl::PointCloud<pcl::PointXYZ>);
pcl::PointCloud<pcl::PointXYZ>::Ptr target(new pcl::PointCloud<pcl::PointXYZ>);
if (pcl::io::loadPCDFile("source.pcd", *source) < 0 ||
pcl::io::loadPCDFile("target.pcd", *target) < 0)
{
pcl::console::print_error("Error loading object/scene file!\n");
return (1);
}
std::vector<int> indices;
pcl::removeNaNFromPointCloud(*source, *source, indices);
pcl::removeNaNFromPointCloud(*target, *target, indices);
//pcl::io::savePCDFileBinary("12300.pcd", *target);
pcl::NormalEstimation<pcl::PointXYZ, pcl::PointNormal> ne;
ne.setInputCloud(source);
pcl::search::KdTree<pcl::PointXYZ>::Ptr tree(new pcl::search::KdTree<pcl::PointXYZ>());
ne.setSearchMethod(tree);
ne.setKSearch(15);
pcl::PointCloud<pcl::PointNormal>::Ptr cloud_normals(new pcl::PointCloud<pcl::PointNormal>);
//ne.setRadiusSearch(2);
ne.compute(*cloud_normals);
pcl::copyPointCloud(*source, *cloud_normals);
//pcl::io::savePCDFileBinary("normal.pcd", *cloud_normals);
//target
pcl::NormalEstimation<pcl::PointXYZ, pcl::PointNormal> ne2;
ne2.setInputCloud(target);
ne2.setSearchMethod(tree);
pcl::PointCloud<pcl::PointNormal>::Ptr cloud_normalt(new pcl::PointCloud<pcl::PointNormal>);
//ne.setRadiusSearch(2);
ne2.setKSearch(15);
ne2.compute(*cloud_normalt);
pcl::copyPointCloud(*target, *cloud_normalt);
// Downsample
//std::cout << object->size() << endl;
//std::cout << scene->size() << endl;
pcl::console::print_highlight("Downsampling...\n");
pcl::PointCloud<pcl::PointNormal>::Ptr cloud_normals_down(new pcl::PointCloud<pcl::PointNormal>);
pcl::PointCloud<pcl::PointNormal>::Ptr cloud_normalt_down(new pcl::PointCloud<pcl::PointNormal>);
pcl::VoxelGrid<pcl::PointNormal> grid;
const float leaf = 0.8f;//0.005f;
grid.setLeafSize(leaf, leaf, leaf);
grid.setInputCloud(cloud_normals);
grid.filter(*cloud_normals_down);
grid.setInputCloud(cloud_normalt);
grid.filter(*cloud_normalt_down);
//pcl::io::savePCDFile("a0.pcd", *cloud_normalt);
//pcl::io::savePCDFile("a.pcd", *cloud_normalt_down);
// Estimate features
pcl::console::print_highlight("Estimating features...\n");
pcl::FPFHEstimation<pcl::PointNormal, pcl::PointNormal, pcl::FPFHSignature33> fpfh;
//pcl::PointCloud<pcl::PointXYZ>::Ptr cloud_normals_xyz(new pcl::PointCloud<pcl::PointXYZ>);
fpfh.setInputCloud(cloud_normals_down);
fpfh.setInputNormals(cloud_normals_down);
//fpfh.setSearchMethod(tree);
pcl::PointCloud<pcl::FPFHSignature33>::Ptr fpfhs(new pcl::PointCloud<pcl::FPFHSignature33>());
pcl::PointCloud<pcl::FPFHSignature33>::Ptr fpfht(new pcl::PointCloud<pcl::FPFHSignature33>());
fpfh.setRadiusSearch(3.5);
fpfh.compute(*fpfhs);
//pfh.setKSearch(30);
// Compute the target features
fpfh.setInputCloud(cloud_normalt_down);
fpfh.setInputNormals(cloud_normalt_down);
//fpfh.setSearchMethod(tree);
fpfh.setRadiusSearch(3.5);
fpfh.compute(*fpfht);
// Perform alignment
pcl::console::print_highlight("Starting alignment...\n");
pcl::SampleConsensusPrerejective<PointNT, PointNT, FeatureT> align;
align.setInputSource(cloud_normals_down);
align.setSourceFeatures(fpfhs);
align.setInputTarget(cloud_normalt_down);
align.setTargetFeatures(fpfht);
align.setMaximumIterations(5000); // Number of RANSAC iterations
align.setNumberOfSamples(3);// Number of points to sample for generating/prerejecting a pose
align.setCorrespondenceRandomness(5); // Number of nearest features to use
align.setSimilarityThreshold(0.9f); // Polygonal edge length similarity threshold
align.setMaxCorrespondenceDistance(1.5f); // Inlier threshold 2.5f* leaf
align.setInlierFraction(0.5f); // Required inlier fraction for accepting a pose hypothesis
{
//pcl::ScopeTime t("Alignment");
align.align(*object_aligned);
}
cout << "计算完毕" << endl;
if (align.hasConverged())
{
// Print results
printf("\n");
Eigen::Matrix4f transformation = align.getFinalTransformation();
pcl::console::print_info(" | %6.3f %6.3f %6.3f | \n", transformation(0, 0), transformation(0, 1), transformation(0, 2));
pcl::console::print_info("R = | %6.3f %6.3f %6.3f | \n", transformation(1, 0), transformation(1, 1), transformation(1, 2));
pcl::console::print_info(" | %6.3f %6.3f %6.3f | \n", transformation(2, 0), transformation(2, 1), transformation(2, 2));
pcl::console::print_info("\n");
pcl::console::print_info("t = < %0.3f, %0.3f, %0.3f >\n", transformation(0, 3), transformation(1, 3), transformation(2, 3));
pcl::console::print_info("\n");
pcl::console::print_info("Inliers: %i/%i\n", align.getInliers().size(), object->size());
pcl::transformPointCloud(*object_aligned, *object_aligned, transformation);
pcl::io::savePCDFile("11111111.pcd", *object_aligned);
// Show alignment
pcl::visualization::PCLVisualizer visu("Alignment");
visu.addPointCloud(scene, ColorHandlerT(scene, 0.0, 255.0, 0.0), "scene");
visu.addPointCloud(object_aligned, ColorHandlerT(object_aligned, 0.0, 0.0, 255.0), "object_aligned");
visu.spin();
}
else
{
pcl::console::print_error("Alignment failed!\n");
return (1);
}
return (0);
}调试参数并不理想
随机函数pcl::SampleConsensusInitialAlignment<pcl::PointXYZ, pcl::PointXYZ, pcl::FPFHSignature33> scia;
#include <pcl/registration/ia_ransac.h>
#include <pcl/point_types.h>
#include <pcl/point_cloud.h>
#include <pcl/features/normal_3d.h>
#include <pcl/features/fpfh.h>
#include <pcl/search/kdtree.h>
#include <pcl/io/pcd_io.h>
#include <pcl/filters/voxel_grid.h>
#include <pcl/filters/filter.h>
#include <pcl/registration/icp.h>
#include <pcl/visualization/pcl_visualizer.h>
#include <time.h>
using pcl::NormalEstimation;
using pcl::search::KdTree;
typedef pcl::PointXYZ PointT;
typedef pcl::PointCloud<PointT> PointCloud;
//点云可视化
void visualize_pcd(PointCloud::Ptr pcd_src,
PointCloud::Ptr pcd_tgt,
PointCloud::Ptr pcd_final)
{
//int vp_1, vp_2;
// Create a PCLVisualizer object
pcl::visualization::PCLVisualizer viewer("registration Viewer");
viewer.initCameraParameters();
//viewer.createViewPort (0.0, 0, 0.5, 1.0, vp_1);
// viewer.createViewPort (0.5, 0, 1.0, 1.0, vp_2);
pcl::visualization::PointCloudColorHandlerCustom<pcl::PointXYZ> src_h(pcd_src, 0, 255, 0);
pcl::visualization::PointCloudColorHandlerCustom<pcl::PointXYZ> tgt_h(pcd_tgt, 255, 0, 0);
pcl::visualization::PointCloudColorHandlerCustom<pcl::PointXYZ> final_h(pcd_final, 0, 0, 255);
viewer.addPointCloud(pcd_src, src_h, "source cloud");
viewer.addPointCloud(pcd_tgt, tgt_h, "tgt cloud");
viewer.addPointCloud(pcd_final, final_h, "final cloud");
//viewer.addCoordinateSystem(1.0);
while (!viewer.wasStopped())
{
viewer.spinOnce(100);
}
}
//由旋转平移矩阵计算旋转角度
void matrix2angle(Eigen::Matrix4f& result_trans, Eigen::Vector3f& result_angle)
{
double ax, ay, az;
if (result_trans(2, 0) == 1 || result_trans(2, 0) == -1)
{
az = 0;
double dlta;
dlta = atan2(result_trans(0, 1), result_trans(0, 2));
if (result_trans(2, 0) == -1)
{
ay = M_PI / 2;
ax = az + dlta;
}
else
{
ay = -M_PI / 2;
ax = -az + dlta;
}
}
else
{
ay = -asin(result_trans(2, 0));
ax = atan2(result_trans(2, 1) / cos(ay), result_trans(2, 2) / cos(ay));
az = atan2(result_trans(1, 0) / cos(ay), result_trans(0, 0) / cos(ay));
}
result_angle << ax, ay, az;
}
int
main(int argc, char** argv)
{
//加载点云文件
PointCloud::Ptr cloud_src_o(new PointCloud);//原点云，待配准
pcl::io::loadPCDFile("source.pcd", *cloud_src_o);
PointCloud::Ptr cloud_tgt_o(new PointCloud);//目标点云
pcl::io::loadPCDFile("target.pcd", *cloud_tgt_o);
clock_t start = clock();
//去除NAN点
std::vector<int> indices_src; //保存去除的点的索引
pcl::removeNaNFromPointCloud(*cloud_src_o, *cloud_src_o, indices_src);
pcl::removeNaNFromPointCloud(*cloud_tgt_o, *cloud_tgt_o, indices_src);
std::cout << "remove *cloud_src_o nan" << endl;
//下采样滤波
pcl::VoxelGrid<pcl::PointXYZ> voxel_grid;
voxel_grid.setLeafSize(0.8, 0.8, 0.8);
voxel_grid.setInputCloud(cloud_src_o);
PointCloud::Ptr cloud_src(new PointCloud);
voxel_grid.filter(*cloud_src);
std::cout << "down size *cloud_src_o from " << cloud_src_o->size() << "to" << cloud_src->size() << endl;
//pcl::io::savePCDFileASCII("bunny_src_down.pcd", *cloud_src);
//计算表面法线
pcl::NormalEstimation<pcl::PointXYZ, pcl::Normal> ne_src;
ne_src.setInputCloud(cloud_src);
pcl::search::KdTree< pcl::PointXYZ>::Ptr tree_src(new pcl::search::KdTree< pcl::PointXYZ>());
ne_src.setSearchMethod(tree_src);
pcl::PointCloud<pcl::Normal>::Ptr cloud_src_normals(new pcl::PointCloud< pcl::Normal>);
ne_src.setRadiusSearch(1.25);
ne_src.compute(*cloud_src_normals);
pcl::VoxelGrid<pcl::PointXYZ> voxel_grid_2;
voxel_grid_2.setLeafSize(0.8, 0.8, 0.8);
voxel_grid_2.setInputCloud(cloud_tgt_o);
PointCloud::Ptr cloud_tgt(new PointCloud);
voxel_grid_2.filter(*cloud_tgt);
std::cout << "down size *cloud_tgt_o.pcd from " << cloud_tgt_o->size() << "to" << cloud_tgt->size() << endl;
//pcl::io::savePCDFileASCII("bunny_tgt_down.pcd", *cloud_tgt);
pcl::NormalEstimation<pcl::PointXYZ, pcl::Normal> ne_tgt;
ne_tgt.setInputCloud(cloud_tgt);
pcl::search::KdTree< pcl::PointXYZ>::Ptr tree_tgt(new pcl::search::KdTree< pcl::PointXYZ>());
ne_tgt.setSearchMethod(tree_tgt);
pcl::PointCloud<pcl::Normal>::Ptr cloud_tgt_normals(new pcl::PointCloud< pcl::Normal>);
//ne_tgt.setKSearch(20);
ne_tgt.setRadiusSearch(1.25);
ne_tgt.compute(*cloud_tgt_normals);
//计算FPFH
pcl::FPFHEstimation<pcl::PointXYZ, pcl::Normal, pcl::FPFHSignature33> fpfh_src;
fpfh_src.setInputCloud(cloud_src);
fpfh_src.setInputNormals(cloud_src_normals);
pcl::search::KdTree<PointT>::Ptr tree_src_fpfh(new pcl::search::KdTree<PointT>);
fpfh_src.setSearchMethod(tree_src_fpfh);
pcl::PointCloud<pcl::FPFHSignature33>::Ptr fpfhs_src(new pcl::PointCloud<pcl::FPFHSignature33>());
fpfh_src.setRadiusSearch(2.5);
fpfh_src.compute(*fpfhs_src);
std::cout << "compute *cloud_src fpfh" << endl;
pcl::FPFHEstimation<pcl::PointXYZ, pcl::Normal, pcl::FPFHSignature33> fpfh_tgt;
fpfh_tgt.setInputCloud(cloud_tgt);
fpfh_tgt.setInputNormals(cloud_tgt_normals);
pcl::search::KdTree<PointT>::Ptr tree_tgt_fpfh(new pcl::search::KdTree<PointT>);
fpfh_tgt.setSearchMethod(tree_tgt_fpfh);
pcl::PointCloud<pcl::FPFHSignature33>::Ptr fpfhs_tgt(new pcl::PointCloud<pcl::FPFHSignature33>());
fpfh_tgt.setRadiusSearch(2.5);
fpfh_tgt.compute(*fpfhs_tgt);
std::cout << "compute *cloud_tgt fpfh" << endl;
//SAC配准
pcl::SampleConsensusInitialAlignment<pcl::PointXYZ, pcl::PointXYZ, pcl::FPFHSignature33> scia;
scia.setInputSource(cloud_src);
scia.setInputTarget(cloud_tgt);
scia.setSourceFeatures(fpfhs_src);
scia.setTargetFeatures(fpfhs_tgt);
//scia.setMinSampleDistance(0.85);
scia.setRANSACIterations(5000);
scia.setNumberOfSamples(3);
scia.setMaximumIterations(5000);
scia.setCorrespondenceRandomness(5);
scia.setMaxCorrespondenceDistance(2.6);
PointCloud::Ptr sac_result(new PointCloud);
scia.align(*sac_result);
pcl::io::savePCDFile("align123.pcd", *sac_result);
std::cout << "sac has converged:" << scia.hasConverged() << " score: " << scia.getFitnessScore() << endl;
Eigen::Matrix4f sac_trans;
sac_trans = scia.getFinalTransformation();
std::cout << sac_trans << endl;
//pcl::io::savePCDFileASCII("bunny_transformed_sac.pcd", *sac_result);
clock_t sac_time = clock();
//icp配准
//PointCloud::Ptr icp_result(new PointCloud);
//pcl::IterativeClosestPoint<pcl::PointXYZ, pcl::PointXYZ> icp;
//icp.setInputSource(sac_result);
//icp.setInputTarget(cloud_tgt_o);
Set the max correspondence distance to 4cm (e.g., correspondences with higher distances will be ignored)
//icp.setMaxCorrespondenceDistance(0.55);
最大迭代次数
//icp.setMaximumIterations(100);
两次变化矩阵之间的差值
//icp.setTransformationEpsilon(1e-10);
均方误差
//icp.setEuclideanFitnessEpsilon(0.2);
//icp.align(*icp_result);
//pcl::io::savePCDFile("icp.pcd", *icp_result);
//clock_t end = clock();
//cout << "total time: " << (double)(end - start) / (double)CLOCKS_PER_SEC << " s" << endl;
把计算法线和点特征直方图的时间也算在SAC里面了
//cout << "sac time: " << (double)(sac_time - start) / (double)CLOCKS_PER_SEC << " s" << endl;
//cout << "icp time: " << (double)(end - sac_time) / (double)CLOCKS_PER_SEC << " s" << endl;
//std::cout << "ICP has converged:" << icp.hasConverged()
// << " score: " << icp.getFitnessScore() << std::endl;
//Eigen::Matrix4f icp_trans;
//icp_trans = icp.getFinalTransformation();
cout<<"ransformationProbability"<<icp.getTransformationProbability()<<endl;
//std::cout << icp_trans << endl;
使用创建的变换对未过滤的输入点云进行变换
//PointCloud::Ptr icptrans_result(new PointCloud);
//pcl::transformPointCloud(*cloud_src_o, *icptrans_result, icp_trans);
保存转换的输入点云
pcl::io::savePCDFileASCII("bunny_transformed_sac_ndt.pcd", *icp_result);
计算误差
//Eigen::Vector3f ANGLE_origin;
//ANGLE_origin << 0, 0, M_PI / 5;
//double error_x, error_y, error_z;
//Eigen::Vector3f ANGLE_result;
//matrix2angle(icp_trans, ANGLE_result);
//error_x = fabs(ANGLE_result(0)) - fabs(ANGLE_origin(0));
//error_y = fabs(ANGLE_result(1)) - fabs(ANGLE_origin(1));
//error_z = fabs(ANGLE_result(2)) - fabs(ANGLE_origin(2));
//cout << "original angle in x y z:\n" << ANGLE_origin << endl;
//cout << "error in aixs_x: " << error_x << " error in aixs_y: " << error_y << " error in aixs_z: " << error_z << endl;
可视化*/
//visualize_pcd(cloud_src_o, cloud_tgt_o, icptrans_result);
return (0);
}
