Exemplo n.º 1
0
void PoissonReconstruction::perform(int ksearch)
{
    PointCloud<PointXYZ>::Ptr cloud (new PointCloud<PointXYZ>);
    PointCloud<PointNormal>::Ptr cloud_with_normals (new PointCloud<PointNormal>);
    search::KdTree<PointXYZ>::Ptr tree;
    search::KdTree<PointNormal>::Ptr tree2;

    cloud->reserve(myPoints.size());
    for (Points::PointKernel::const_iterator it = myPoints.begin(); it != myPoints.end(); ++it) {
        if (!boost::math::isnan(it->x) && !boost::math::isnan(it->y) && !boost::math::isnan(it->z))
            cloud->push_back(PointXYZ(it->x, it->y, it->z));
    }

    // Create search tree
    tree.reset (new search::KdTree<PointXYZ> (false));
    tree->setInputCloud (cloud);

    // Normal estimation
    NormalEstimation<PointXYZ, Normal> n;
    PointCloud<Normal>::Ptr normals (new PointCloud<Normal> ());
    n.setInputCloud (cloud);
    //n.setIndices (indices[B);
    n.setSearchMethod (tree);
    n.setKSearch (ksearch);
    n.compute (*normals);

    // Concatenate XYZ and normal information
    pcl::concatenateFields (*cloud, *normals, *cloud_with_normals);

    // Create search tree
    tree2.reset (new search::KdTree<PointNormal>);
    tree2->setInputCloud (cloud_with_normals);

    // Init objects
    Poisson<PointNormal> poisson;

    // Set parameters
    poisson.setInputCloud (cloud_with_normals);
    poisson.setSearchMethod (tree2);
    if (depth >= 1)
        poisson.setDepth(depth);
    if (solverDivide >= 1)
        poisson.setSolverDivide(solverDivide);
    if (samplesPerNode >= 1.0f)
        poisson.setSamplesPerNode(samplesPerNode);

    // Reconstruct
    PolygonMesh mesh;
    poisson.reconstruct (mesh);

    MeshConversion::convert(mesh, myMesh);
}
Exemplo n.º 2
0
void PoissonReconstruction::perform(const std::vector<Base::Vector3f>& normals)
{
    if (myPoints.size() != normals.size())
        throw Base::RuntimeError("Number of points doesn't match with number of normals");

    PointCloud<PointNormal>::Ptr cloud_with_normals (new PointCloud<PointNormal>);
    search::KdTree<PointNormal>::Ptr tree;

    cloud_with_normals->reserve(myPoints.size());
    std::size_t num_points = myPoints.size();
    const std::vector<Base::Vector3f>& points = myPoints.getBasicPoints();
    for (std::size_t index=0; index<num_points; index++) {
        const Base::Vector3f& p = points[index];
        const Base::Vector3f& n = normals[index];
        if (!boost::math::isnan(p.x) && !boost::math::isnan(p.y) && !boost::math::isnan(p.z)) {
            PointNormal pn;
            pn.x = p.x;
            pn.y = p.y;
            pn.z = p.z;
            pn.normal_x = n.x;
            pn.normal_y = n.y;
            pn.normal_z = n.z;
            cloud_with_normals->push_back(pn);
        }
    }

    // Create search tree
    tree.reset (new search::KdTree<PointNormal>);
    tree->setInputCloud (cloud_with_normals);

    // Init objects
    Poisson<PointNormal> poisson;

    // Set parameters
    poisson.setInputCloud (cloud_with_normals);
    poisson.setSearchMethod (tree);
    if (depth >= 1)
        poisson.setDepth(depth);
    if (solverDivide >= 1)
        poisson.setSolverDivide(solverDivide);
    if (samplesPerNode >= 1.0f)
        poisson.setSamplesPerNode(samplesPerNode);

    // Reconstruct
    PolygonMesh mesh;
    poisson.reconstruct (mesh);

    MeshConversion::convert(mesh, myMesh);
}