首先将点云分块得到区域后，获取每个块的box的最大最小点云，然后提取box内的点云。
pcl::IndicesPtr indexes(new pcl::Indices());
pcl::getPointsInBox(*cloud_1, min_pt, max_pt, *indexes);
// --------------------------取框内和框外点------------------------------------
pcl::ExtractIndices<PointType> extr;
extr.setInputCloud(cloud); // 设置输入点云
extr.setIndices(indexes); // 设置索引
pcl::PointCloud<PointType>::Ptr cloud_in_box(new pcl::PointCloud<PointType>());
extr.filter(*cloud_in_box); // 提取对应索引的点云
cout << "方框内点的个数为：" << cloud_in_box->points.size() << endl;
pcl::PointCloud<PointType>::Ptr cloud_out_box(new pcl::PointCloud<PointType>);
extr.setNegative(true); // 提取对应索引之外的点
extr.filter(*cloud_out_box);
cout << "方框外点的个数为：" << cloud_out_box->points.size() << endl;但是不知道为什么每次提取的indexes都不对，提取的区域大部分都是包含整个点云的点，很奇怪。于是看了下提取box点云的源码
pcl::IndicesPtr indexes(new pcl::Indices());
pcl::getPointsInBox(*cloud_1, min_pt, max_pt, *indexes);//
template <typename PointT> inline void
pcl::getPointsInBox (const pcl::PointCloud<PointT> &cloud,
Eigen::Vector4f &min_pt, Eigen::Vector4f &max_pt,
Indices &indices)
{
indices.resize (cloud.size ());
int l = 0;
// If the data is dense, we don't need to check for NaN
if (cloud.is_dense)
{
for (std::size_t i = 0; i < cloud.size (); ++i)
{
// Check if the point is inside bounds
if (cloud[i].x < min_pt[0] || cloud[i].y < min_pt[1] || cloud[i].z < min_pt[2])
continue;
if (cloud[i].x > max_pt[0] || cloud[i].y > max_pt[1] || cloud[i].z > max_pt[2])
continue;
indices[l++] = int (i);
}
}
// NaN or Inf values could exist => check for them
else
{
for (std::size_t i = 0; i < cloud.size (); ++i)
{
// Check if the point is invalid
if (!std::isfinite (cloud[i].x) ||
!std::isfinite (cloud[i].y) ||
!std::isfinite (cloud[i].z))
continue;
// Check if the point is inside bounds
if (cloud[i].x < min_pt[0] || cloud[i].y < min_pt[1] || cloud[i].z < min_pt[2])
continue;
if (cloud[i].x > max_pt[0] || cloud[i].y > max_pt[1] || cloud[i].z > max_pt[2])
continue;
indices[l++] = int (i);
}
}
indices.resize (l);
}
不知道为什么不行，但看到只是比价点的大小是否在box内，按照源码修改，完美解决。
//
// Created by wzs on 2024/6/6.
//
#include <vector>
#include <string>
#include <iostream>
#include <sstream>
#include <stdlib.h> //rand()头文件
#include <pcl/io/pcd_io.h>
#include <pcl/octree/octree.h>
#include <boost/thread/thread.hpp>
#include <pcl/visualization/pcl_visualizer.h>
#include <pcl/filters/extract_indices.h>
#include <pcl/common/common.h>
#include "pcl/common/centroid.h"
#include "pcl/common/distances.h"
using namespace std;
typedef pcl::PointXYZ PointType;
int main()
{
string path("../test_data/1/3.pcd");// 原始点云所在文件夹
string outpath("../test_data/1/voxel_map"); // 分块保存路径文件夹名
string path_1("../test_data/1/3_slope.pcd");
string outpath_1("../test_data/1/voxel_map_1"); // 分块保存路径文件夹名
string outpath_2("../test_data/1/voxel_map_2"); // 分块保存路径文件夹名
float resolution = 1.0; // 设置体素分辨率
//-------------------------- 加载点云 --------------------------
pcl::PointCloud<PointType>::Ptr cloud(new pcl::PointCloud<PointType>);
if (pcl::io::loadPCDFile(path, *cloud) < 0)
{
PCL_ERROR("Couldn't read file \n");
// system("pause");
return -1;
}
pcl::PointCloud<PointType>::Ptr cloud_1(new pcl::PointCloud<PointType>);
if (pcl::io::loadPCDFile(path_1, *cloud_1) < 0)
{
PCL_ERROR("cloud_slopen't read file \n");
// system("pause");
return -1;
}
// -----------------------获取分块点云保存路径----------------------------------
printf("开始进行点云分块！！\n");
// string::size_type iPos = path.find_last_of('//') + 1;
// string filename = path.substr(iPos, path.length() - iPos);
// string name = filename.substr(0, filename.rfind("."));
// ---------------------使用八叉树快速构建体素索引------------------------------
pcl::octree::OctreePointCloud<PointType> octree(resolution);
octree.setInputCloud(cloud);
octree.addPointsFromInputCloud(); // 从点云中构建八叉树
pcl::Indices pointIdxVec; // 体素内点的索引
//-----------------------------开始分块-----------------------------------------
// boost::shared_ptr<pcl::visualization::PCLVisualizer>viewer(new pcl::visualization::PCLVisualizer("planar segment")); ;
// viewer->setBackgroundColor(0, 0, 0);
// viewer->setWindowName("点云分块");
pcl::PointCloud<PointType>::Ptr seg_cloud(new pcl::PointCloud<PointType>);
// pcl::PointCloud<PointType>::Ptr seg_cloud_1(new pcl::PointCloud<PointType>);
// pcl::PointCloud<PointType>::Ptr seg_cloud_1_0(new pcl::PointCloud<PointType>);
int num_voxel = 0;
for (auto iter = octree.leaf_breadth_begin(); iter != octree.leaf_breadth_end(); ++iter){
num_voxel++;
}
float mean_voxel_size_ = cloud_1->size() / num_voxel;
int begin = 0;
// 构建八叉树叶子节点迭代器，遍历八叉树
for (auto iter = octree.leaf_breadth_begin(); iter != octree.leaf_breadth_end(); ++iter)
{
auto key = iter.getCurrentOctreeKey(); // 获取当前迭代器八叉树节点的八叉树键
auto it_key = octree.findLeaf(key.x, key.y, key.z); // 检查是否存在于八叉树中
if (it_key != nullptr)
{
pointIdxVec.clear();
//从八叉树叶子节点中获取单个叶子容器中的点索引
pointIdxVec = iter.getLeafContainer().getPointIndicesVector();
if (pointIdxVec.size() == 0) // 体素内点个数为0则跳过
{
continue;
}else
{
seg_cloud->clear();
std::stringstream ss;
ss << outpath << "//"<< "_block_" << begin + 1 << ".pcd";
pcl::copyPointCloud(*cloud, pointIdxVec, *seg_cloud);
pcl::io::savePCDFileBinary(ss.str(), *seg_cloud);
cout << "第[" << begin + 1 << "]块点云分割完毕！ " << seg_cloud->size() << " " << mean_voxel_size_ << endl;
pcl::PointCloud<PointType>::Ptr cloud_in_box(new pcl::PointCloud<PointType>());
PointType point_min_, point_max_;
pcl::getMinMax3D(*seg_cloud,point_min_,point_max_);
// cout << "Min x: " << point_min_.x << endl;
// cout << "Min y: " << point_min_.y << endl;
// cout << "Min z: " << point_min_.z << endl;
// cout << "Max x: " << point_max_.x << endl;
// cout << "Max y: " << point_max_.y << endl;
// cout << "Max z: " << point_max_.z << endl;
// Eigen::Vector4f min_pt = { point_min_.x,point_min_.y,point_min_.z,0 };
// Eigen::Vector4f max_pt = { point_max_.x,point_max_.y,point_max_.z,0 };
for (std::size_t i = 0; i < cloud_1->size (); ++i)
{
// Check if the point is inside bounds
if (cloud_1->points[i].x < point_min_.x || cloud_1->points[i].y < point_min_.y || cloud_1->points[i].z < point_min_.y)
continue;
if (cloud_1->points[i].x > point_max_.x || cloud_1->points[i].y > point_max_.y || cloud_1->points[i].z > point_max_.z)
continue;
cloud_in_box->emplace_back(cloud_1->points[i]);
}
cout << "方框内点的个数为：" << cloud_in_box->points.size() << endl;
std::stringstream ss_1;
ss_1 << outpath_1 << "//"<< "_block_" << begin + 1 << ".pcd";
if(cloud_in_box->size() > 0)
pcl::io::savePCDFileBinary(ss_1.str(), *cloud_in_box);
//center distance
Eigen::Vector4f centroid_seg_, centroid_i_box_;
PointType centroid_seg_p_, centroid_i_box_p_;
pcl::compute3DCentroid(*seg_cloud,centroid_seg_);
pcl::compute3DCentroid(*cloud_in_box,centroid_i_box_);
centroid_seg_p_.x = centroid_seg_[0];
centroid_seg_p_.y = centroid_seg_[1];
centroid_seg_p_.z = centroid_seg_[2];
centroid_i_box_p_.x = centroid_i_box_[0];
centroid_i_box_p_.y = centroid_i_box_[1];
centroid_i_box_p_.z = centroid_i_box_[2];
float dis_centroid_seg_in_box_ = pcl::euclideanDistance(centroid_seg_p_,centroid_i_box_p_);
cout << "dis_centroid_seg_in_box_: " << dis_centroid_seg_in_box_ << endl;
std::stringstream ss_2;
ss_2 << outpath_2 << "//"<< "_block_" << begin + 1 << ".pcd";
if (dis_centroid_seg_in_box_ > 0.1 && cloud_in_box->size() > mean_voxel_size_){
pcl::io::savePCDFileBinary(ss_2.str(), *cloud_in_box);
cout << begin + 1 << endl;
}
begin++;
}
seg_cloud->clear(); // 每分割一次，清空一下容器，进而提高电脑工作效率
}
}
printf("点云体素分块完毕！！！");
return 0;
}
效果图如下：
