mirror of https://github.com/CGAL/cgal
New algorithm to smooth 3D point cloud inspired y Improved Laplacian smoothing of noisy surface meshes
This commit is contained in:
parent
ee527de50f
commit
9fe8881ed7
|
|
@ -0,0 +1,430 @@
|
|||
// Copyright (c) 2007-09 INRIA Sophia-Antipolis (France).
|
||||
// All rights reserved.
|
||||
//
|
||||
// This file is part of CGAL (www.cgal.org); you may redistribute it under
|
||||
// the terms of the Q Public License version 1.0.
|
||||
// See the file LICENSE.QPL distributed with CGAL.
|
||||
//
|
||||
// Licensees holding a valid commercial license may use this file in
|
||||
// accordance with the commercial license agreement provided with the software.
|
||||
//
|
||||
// This file is provided AS IS with NO WARRANTY OF ANY KIND, INCLUDING THE
|
||||
// WARRANTY OF DESIGN, MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE.
|
||||
//
|
||||
// $URL$
|
||||
// $Id$
|
||||
//
|
||||
// Author(s) : Nader Salman and Laurent Saboret
|
||||
|
||||
#ifndef CGAL_IMPROVED_LAPLACIAN_SMOOTHING_3_H
|
||||
#define CGAL_IMPROVED_LAPLACIAN_SMOOTHING_3_H
|
||||
|
||||
#include <CGAL/Search_traits_3.h>
|
||||
#include <CGAL/Orthogonal_k_neighbor_search.h>
|
||||
#include <iterator>
|
||||
#include <list>
|
||||
|
||||
CGAL_BEGIN_NAMESPACE
|
||||
|
||||
|
||||
// ----------------------------------------------------------------------------
|
||||
// Private section
|
||||
// ----------------------------------------------------------------------------
|
||||
namespace CGALi {
|
||||
|
||||
|
||||
// Item in the Kd-tree: position (Point_3) + index
|
||||
template <typename Kernel>
|
||||
class KdTreeElement : public Kernel::Point_3
|
||||
{
|
||||
public:
|
||||
unsigned int index;
|
||||
|
||||
// basic geometric types
|
||||
typedef typename CGAL::Origin Origin;
|
||||
typedef typename Kernel::Point_3 Point_3;
|
||||
|
||||
KdTreeElement(const Origin& o = ORIGIN, unsigned int id=0)
|
||||
: Point_3(o), index(id)
|
||||
{}
|
||||
KdTreeElement(const Point_3& p, unsigned int id=0)
|
||||
: Point_3(p), index(id)
|
||||
{}
|
||||
KdTreeElement(const KdTreeElement& other)
|
||||
: Point_3(other), index(other.index)
|
||||
{}
|
||||
};
|
||||
|
||||
// Helper class for the Kd-tree
|
||||
template <typename Kernel>
|
||||
class KdTreeGT : public Kernel
|
||||
{
|
||||
public:
|
||||
typedef KdTreeElement<Kernel> Point_3;
|
||||
};
|
||||
|
||||
template <typename Kernel>
|
||||
class KdTreeTraits : public CGAL::Search_traits_3<KdTreeGT<Kernel> >
|
||||
{
|
||||
public:
|
||||
typedef typename Kernel::Point_3 PointType;
|
||||
};
|
||||
|
||||
/* Usage:
|
||||
typedef CGALi::KdTreeElement<Kernel> KdTreeElement;
|
||||
typedef CGALi::KdTreeTraits<Kernel> Tree_traits;
|
||||
typedef CGAL::Orthogonal_k_neighbor_search<Tree_traits> Neighbor_search;
|
||||
typedef typename Neighbor_search::Tree Tree;
|
||||
typedef typename Neighbor_search::iterator Search_iterator;
|
||||
*/
|
||||
|
||||
|
||||
/// Smooth one point position using jet fitting on the k
|
||||
/// nearest neighbors and reprojection onto the jet.
|
||||
///
|
||||
/// @commentheading Precondition: k >= 2.
|
||||
///
|
||||
/// @commentheading Template Parameters:
|
||||
/// @param Kernel Geometric traits class.
|
||||
/// @param Tree KD-tree.
|
||||
///
|
||||
/// @return computed point
|
||||
template <typename Kernel,
|
||||
typename Tree>
|
||||
typename Kernel::Point_3
|
||||
laplacian_smoothing_3(
|
||||
const typename Kernel::Point_3& pi, ///< 3D point to smooth
|
||||
Tree& tree, ///< KD-tree
|
||||
const unsigned int k)
|
||||
{
|
||||
// basic geometric types
|
||||
typedef typename Kernel::Point_3 Point_3;
|
||||
typedef typename Kernel::Vector_3 Vector_3;
|
||||
|
||||
// types for K nearest neighbors search
|
||||
//typedef typename CGAL::Search_traits_3<Kernel> Tree_traits;
|
||||
typedef CGALi::KdTreeTraits<Kernel> Tree_traits;
|
||||
typedef CGAL::Orthogonal_k_neighbor_search<Tree_traits> Neighbor_search;
|
||||
typedef typename Neighbor_search::iterator Search_iterator;
|
||||
|
||||
// Compute Laplacian (centroid) of k neighboring points.
|
||||
// Note: we perform k+1 queries and skip the query point which is output first.
|
||||
// TODO: we should use the functions in PCA component instead.
|
||||
Vector_3 v = CGAL::NULL_VECTOR;
|
||||
Neighbor_search search(tree,pi,k+1);
|
||||
Search_iterator search_iterator;
|
||||
for(search_iterator = search.begin(), search_iterator++; // skip pi point
|
||||
search_iterator != search.end();
|
||||
search_iterator++ )
|
||||
{
|
||||
const Point_3& p = search_iterator->first;
|
||||
v = v + (p - CGAL::ORIGIN);
|
||||
}
|
||||
|
||||
Point_3 centroid = CGAL::ORIGIN + v / k;
|
||||
|
||||
return centroid;
|
||||
}
|
||||
|
||||
|
||||
/// Smooth one point position using jet fitting on the k
|
||||
/// nearest neighbors and reprojection onto the jet.
|
||||
///
|
||||
/// @commentheading Precondition: k >= 2.
|
||||
///
|
||||
/// @commentheading Template Parameters:
|
||||
/// @param Kernel Geometric traits class.
|
||||
/// @param Tree KD-tree.
|
||||
///
|
||||
/// @return computed point
|
||||
template <typename Kernel,
|
||||
typename Tree>
|
||||
typename Kernel::Point_3
|
||||
improved_laplacian_smoothing_3(
|
||||
const typename Kernel::Point_3& pi, ///< 3D point to smooth
|
||||
const typename Kernel::Vector_3& bi, ///< bi movement
|
||||
Tree& tree, ///< KD-tree
|
||||
const std::vector<typename Kernel::Vector_3>& b,
|
||||
const unsigned int k,
|
||||
typename Kernel::FT beta)
|
||||
{
|
||||
// basic geometric types
|
||||
typedef typename Kernel::Point_3 Point_3;
|
||||
typedef typename Kernel::Vector_3 Vector_3;
|
||||
|
||||
// types for K nearest neighbors search
|
||||
//typedef typename CGAL::Search_traits_3<Kernel> Tree_traits;
|
||||
typedef CGALi::KdTreeTraits<Kernel> Tree_traits;
|
||||
typedef CGAL::Orthogonal_k_neighbor_search<Tree_traits> Neighbor_search;
|
||||
typedef typename Neighbor_search::iterator Search_iterator;
|
||||
|
||||
// Gather set of k neighboring points and compute the sum of their b[] values.
|
||||
// Note: we perform k+1 queries and skip the query point which is output first.
|
||||
Vector_3 bj_sum;
|
||||
Neighbor_search search(tree,pi,k+1);
|
||||
Search_iterator search_iterator;
|
||||
for(search_iterator = search.begin(), search_iterator++; // skip pi point
|
||||
search_iterator != search.end();
|
||||
search_iterator++ )
|
||||
{
|
||||
bj_sum = bj_sum + b[search_iterator->first.index];
|
||||
}
|
||||
|
||||
return pi - (beta * bi + ((1-beta)/k)*bj_sum);
|
||||
}
|
||||
|
||||
|
||||
} /* namespace CGALi */
|
||||
|
||||
|
||||
// ----------------------------------------------------------------------------
|
||||
// Public section
|
||||
// ----------------------------------------------------------------------------
|
||||
|
||||
|
||||
/// Improved Laplacian smoothing (Vollmer et al)
|
||||
/// on the k nearest neighbors.
|
||||
/// This variant requires the kernel.
|
||||
///
|
||||
/// @commentheading Precondition: k >= 2.
|
||||
///
|
||||
/// @commentheading Template Parameters:
|
||||
/// @param InputIterator value_type convertible to Point_3.
|
||||
/// @param OutputIterator value_type convertible to Point_3.
|
||||
/// @param Kernel Geometric traits class.
|
||||
///
|
||||
/// @return past-the-end output iterator.
|
||||
template <typename InputIterator,
|
||||
typename OutputIterator,
|
||||
typename Kernel
|
||||
>
|
||||
OutputIterator
|
||||
improved_laplacian_smoothing_3(
|
||||
InputIterator first, ///< iterator over the first input point
|
||||
InputIterator beyond, ///< past-the-end iterator over input points
|
||||
OutputIterator output, ///< iterator over the first output point
|
||||
const unsigned int k, ///< number of neighbors
|
||||
const unsigned int iter_number,
|
||||
const Kernel& /*kernel*/,
|
||||
typename Kernel::FT alpha,
|
||||
typename Kernel::FT beta)
|
||||
{
|
||||
// Point_3 types
|
||||
typedef typename std::iterator_traits<InputIterator>::value_type Input_point_3;
|
||||
typedef typename Kernel::Point_3 Point_3;
|
||||
typedef typename Kernel::Vector_3 Vector_3;
|
||||
|
||||
// types for K nearest neighbors search structure
|
||||
//typedef typename CGAL::Search_traits_3<Kernel> Tree_traits;
|
||||
typedef CGALi::KdTreeElement<Kernel> KdTreeElement;
|
||||
typedef CGALi::KdTreeTraits<Kernel> Tree_traits;
|
||||
typedef CGAL::Orthogonal_k_neighbor_search<Tree_traits> Neighbor_search;
|
||||
typedef typename Neighbor_search::Tree Tree;
|
||||
typedef typename Neighbor_search::iterator Search_iterator;
|
||||
|
||||
// precondition: at least one element in the container.
|
||||
// to fix: should have at least three distinct points
|
||||
// but this is costly to check
|
||||
CGAL_precondition(first != beyond);
|
||||
|
||||
// precondition: at least 2 nearest neighbors
|
||||
CGAL_precondition(k >= 2);
|
||||
|
||||
unsigned int i; // point index
|
||||
|
||||
// Create kd-tree
|
||||
//Tree tree(first,beyond);
|
||||
std::vector<KdTreeElement> treeElements;
|
||||
for (InputIterator it = first, i=0 ; it != beyond ; ++it,++i)
|
||||
{
|
||||
treeElements.push_back(KdTreeElement(*it,i));
|
||||
}
|
||||
Tree tree(treeElements.begin(), treeElements.end());
|
||||
|
||||
std::vector<Point_3> p; // positions at step iter_n
|
||||
std::vector<Vector_3> b; // ...
|
||||
|
||||
for(InputIterator it = first, i=0; it != beyond; it++, ++i)
|
||||
p[i] = *it;
|
||||
|
||||
for(int iter_n = 0; iter_n < iter_number ; ++iter_n)
|
||||
{
|
||||
// Iterate over input points, compute (original) Laplacian smooth and b[].
|
||||
for(InputIterator it = first, i=0; it != beyond; it++, ++i)
|
||||
{
|
||||
Point_3 np = CGALi::laplacian_smoothing_3<Kernel>(*it,tree,k);
|
||||
b[i] = alpha*(np - *it) + (1-alpha)*(np - p[i]);
|
||||
p[i] = np;
|
||||
}
|
||||
|
||||
// Iterate over input points, compute and output smooth points.
|
||||
// Note: the cast to (Point_3&) ensures compatibility with classes derived from Point_3.
|
||||
for(InputIterator it = first, i=0; it != beyond; it++, ++i)
|
||||
{
|
||||
p[i] = CGALi::improved_laplacian_smoothing_3<Kernel>(p[i],b[i],tree,b,k,beta);
|
||||
}
|
||||
}
|
||||
|
||||
// Iterate over input points and output smooth points.
|
||||
// Note: the cast to (Point_3&) ensures compatibility with classes derived from Point_3.
|
||||
for(InputIterator it = first, i=0; it != beyond; it++, ++i)
|
||||
{
|
||||
Input_point_3 point = *it;
|
||||
(Point_3&)(point) = p[i];
|
||||
*output++ = point;
|
||||
}
|
||||
|
||||
return output;
|
||||
}
|
||||
|
||||
/// Improved Laplacian smoothing (Vollmer et al)
|
||||
/// on the k nearest neighbors.
|
||||
/// This function is mutating the input point set.
|
||||
/// This variant requires the kernel.
|
||||
///
|
||||
/// Warning:
|
||||
/// This method moves the points, thus
|
||||
/// should not be called on containers sorted wrt points position.
|
||||
///
|
||||
/// @commentheading Precondition: k >= 2.
|
||||
///
|
||||
/// @commentheading Template Parameters:
|
||||
/// @param ForwardIterator value_type convertible to Point_3.
|
||||
/// @param Kernel Geometric traits class.
|
||||
template <typename ForwardIterator,
|
||||
typename Kernel>
|
||||
void
|
||||
improved_laplacian_smoothing_3(
|
||||
ForwardIterator first, ///< iterator over the first input/output point
|
||||
ForwardIterator beyond, ///< past-the-end iterator
|
||||
const unsigned int k, ///< number of neighbors
|
||||
const unsigned int iter_number,
|
||||
const Kernel& /*kernel*/,
|
||||
typename Kernel::FT alpha,
|
||||
typename Kernel::FT beta)
|
||||
{
|
||||
// Point_3 types
|
||||
typedef typename std::iterator_traits<ForwardIterator>::value_type Input_point_3;
|
||||
typedef typename Kernel::Point_3 Point_3;
|
||||
typedef typename Kernel::Vector_3 Vector_3;
|
||||
|
||||
// types for K nearest neighbors search structure
|
||||
//typedef typename CGAL::Search_traits_3<Kernel> Tree_traits;
|
||||
typedef CGALi::KdTreeElement<Kernel> KdTreeElement;
|
||||
typedef CGALi::KdTreeTraits<Kernel> Tree_traits;
|
||||
typedef CGAL::Orthogonal_k_neighbor_search<Tree_traits> Neighbor_search;
|
||||
typedef typename Neighbor_search::Tree Tree;
|
||||
typedef typename Neighbor_search::iterator Search_iterator;
|
||||
|
||||
// precondition: at least one element in the container.
|
||||
// to fix: should have at least three distinct points
|
||||
// but this is costly to check
|
||||
CGAL_precondition(first != beyond);
|
||||
|
||||
// precondition: at least 2 nearest neighbors
|
||||
CGAL_precondition(k >= 2);
|
||||
|
||||
unsigned int i; // point index
|
||||
ForwardIterator it; // point iterator
|
||||
|
||||
// Number of input points
|
||||
int nb_points = std::distance(first, beyond);
|
||||
|
||||
// Create kd-tree
|
||||
//Tree tree(first,beyond);
|
||||
std::vector<KdTreeElement> treeElements;
|
||||
for (it = first, i=0 ; it != beyond ; ++it,++i)
|
||||
{
|
||||
treeElements.push_back(KdTreeElement(*it,i));
|
||||
}
|
||||
Tree tree(treeElements.begin(), treeElements.end());
|
||||
|
||||
std::vector<Point_3> p(nb_points); // positions at step iter_n
|
||||
std::vector<Vector_3> b(nb_points); // ...
|
||||
|
||||
for(it = first, i=0; it != beyond; it++, ++i)
|
||||
p[i] = *it;
|
||||
|
||||
for(int iter_n = 0; iter_n < iter_number ; ++iter_n)
|
||||
{
|
||||
// Iterate over input points, compute (original) Laplacian smooth and b[].
|
||||
for(it = first, i=0; it != beyond; it++, ++i)
|
||||
{
|
||||
Point_3 np = CGALi::laplacian_smoothing_3<Kernel>(*it,tree,k);
|
||||
b[i] = alpha*(np - *it) + (1-alpha)*(np - p[i]);
|
||||
p[i] = np;
|
||||
}
|
||||
|
||||
// Iterate over input points, compute and output smooth points.
|
||||
// Note: the cast to (Point_3&) ensures compatibility with classes derived from Point_3.
|
||||
for(it = first, i=0; it != beyond; it++, ++i)
|
||||
{
|
||||
p[i] = CGALi::improved_laplacian_smoothing_3<Kernel>(p[i],b[i],tree,b,k,beta);
|
||||
}
|
||||
}
|
||||
|
||||
// Iterate over input points and mutate them.
|
||||
// Note: the cast to (Point_3&) ensures compatibility with classes derived from Point_3.
|
||||
for(it = first, i=0; it != beyond; it++, ++i)
|
||||
(Point_3&)(*it) = p[i];
|
||||
}
|
||||
|
||||
|
||||
/// Improved Laplacian smoothing (Vollmer et al)
|
||||
/// on the k nearest neighbors.
|
||||
/// This variant deduces the kernel from iterator types.
|
||||
///
|
||||
/// @commentheading Precondition: k >= 2.
|
||||
///
|
||||
/// @return past-the-end output iterator.
|
||||
template <typename InputIterator,
|
||||
typename OutputIterator
|
||||
>
|
||||
OutputIterator
|
||||
improved_laplacian_smoothing_3(
|
||||
InputIterator first, ///< iterator over the first input point
|
||||
InputIterator beyond, ///< past-the-end iterator over input points
|
||||
OutputIterator output, ///< iterator over the first output point
|
||||
unsigned int k, ///< number of neighbors
|
||||
const unsigned int iter_number,
|
||||
double alpha,
|
||||
double beta)
|
||||
{
|
||||
typedef typename std::iterator_traits<InputIterator>::value_type Input_point_3;
|
||||
typedef typename Kernel_traits<Input_point_3>::Kernel Kernel;
|
||||
return improved_laplacian_smoothing_3(first,beyond,output,k,iter_number,Kernel(),alpha, beta);
|
||||
}
|
||||
|
||||
/// Improved Laplacian smoothing (Vollmer et al)
|
||||
/// on the k nearest neighbors.
|
||||
/// This function is mutating the input point set.
|
||||
/// This variant deduces the kernel from iterator types.
|
||||
///
|
||||
/// Warning:
|
||||
/// As this method relocates the points, it
|
||||
/// should not be called on containers sorted w.r.t. point locations.
|
||||
///
|
||||
/// @commentheading Precondition: k >= 2.
|
||||
///
|
||||
/// @commentheading Template Parameters:
|
||||
/// @param ForwardIterator value_type convertible to Point_3.
|
||||
template <typename ForwardIterator>
|
||||
void
|
||||
improved_laplacian_smoothing_3(
|
||||
ForwardIterator first, ///< iterator over the first input/output point
|
||||
ForwardIterator beyond, ///< past-the-end iterator
|
||||
unsigned int k, ///< number of neighbors
|
||||
const unsigned int iter_number,
|
||||
double alpha,
|
||||
double beta)
|
||||
{
|
||||
typedef typename std::iterator_traits<ForwardIterator>::value_type Input_point_3;
|
||||
typedef typename Kernel_traits<Input_point_3>::Kernel Kernel;
|
||||
improved_laplacian_smoothing_3(first,beyond,k,iter_number,Kernel(),alpha, beta);
|
||||
}
|
||||
|
||||
|
||||
CGAL_END_NAMESPACE
|
||||
|
||||
#endif // CGAL_IMPROVED_LAPLACIAN_SMOOTHING_3_H
|
||||
|
||||
Loading…
Reference in New Issue