Point Cloud Library (PCL) 1.12.1
gpu_extract_labeled_clusters.hpp
1/*
2 * Software License Agreement (BSD License)
3 *
4 * Point Cloud Library (PCL) - www.pointclouds.org
5 * Copyright (c) 2009, Willow Garage, Inc.
6 * All rights reserved.
7 *
8 * Redistribution and use in source and binary forms, with or without
9 * modification, are permitted provided that the following conditions
10 * are met:
11 *
12 * * Redistributions of source code must retain the above copyright
13 * notice, this list of conditions and the following disclaimer.
14 * * Redistributions in binary form must reproduce the above
15 * copyright notice, this list of conditions and the following
16 * disclaimer in the documentation and/or other materials provided
17 * with the distribution.
18 * * Neither the name of Willow Garage, Inc. nor the names of its
19 * contributors may be used to endorse or promote products derived
20 * from this software without specific prior written permission.
21 *
22 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
23 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
24 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
25 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
26 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
27 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
28 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
29 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
30 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
31 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
32 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
33 * POSSIBILITY OF SUCH DAMAGE.
34 *
35 * $Id:$
36 *
37 */
38
39#pragma once
40
41#include <pcl/gpu/segmentation/gpu_extract_labeled_clusters.h>
42
43template <typename PointT>
44void
46 const typename pcl::PointCloud<PointT>::Ptr& host_cloud_,
47 const pcl::gpu::Octree::Ptr& tree,
48 float tolerance,
49 std::vector<PointIndices>& clusters,
50 unsigned int min_pts_per_cluster,
51 unsigned int max_pts_per_cluster)
52{
53
54 // Create a bool vector of processed point indices, and initialize it to false
55 // cloud is a DeviceArray<PointType>
56 std::vector<bool> processed(host_cloud_->size(), false);
57
58 int max_answers;
59
60 if (max_pts_per_cluster > host_cloud_->size())
61 max_answers = static_cast<int>(host_cloud_->size());
62 else
63 max_answers = max_pts_per_cluster;
64
65 // to store the current cluster
67
68 // Process all points in the cloud
69 for (std::size_t i = 0; i < host_cloud_->size(); ++i) {
70 // if we already processed this point continue with the next one
71 if (processed[i])
72 continue;
73 // now we will process this point
74 processed[i] = true;
75
76 // Create the query queue on the device, point based not indices
77 pcl::gpu::Octree::Queries queries_device;
78 // Create the query queue on the host
80
81 // Buffer in a new PointXYZ type
82 PointT t = (*host_cloud_)[i];
83 PointXYZ p;
84 p.x = t.x;
85 p.y = t.y;
86 p.z = t.z;
87
88 // Push the starting point in the vector
89 queries_host.push_back(p);
90 // Clear vector
91 r.indices.clear();
92 // Push the starting point in
93 r.indices.push_back(static_cast<int>(i));
94
95 unsigned int found_points = static_cast<unsigned int>(queries_host.size());
96 unsigned int previous_found_points = 0;
97
98 pcl::gpu::NeighborIndices result_device;
99
100 // once the area stop growing, stop also iterating.
101 while (previous_found_points < found_points) {
102 // Move queries to GPU
103 queries_device.upload(queries_host);
104 // Execute search
105 tree->radiusSearch(queries_device, tolerance, max_answers, result_device);
106
107 // Store the previously found number of points
108 previous_found_points = found_points;
109
110 // Host buffer for results
111 std::vector<int> sizes, data;
112
113 // Copy results from GPU to Host
114 result_device.sizes.download(sizes);
115 result_device.data.download(data);
116
117 for (std::size_t qp = 0; qp < sizes.size(); qp++) {
118 for (int qp_r = 0; qp_r < sizes[qp]; qp_r++) {
119 if (processed[data[qp_r + qp * max_answers]])
120 continue;
121 // Only add if label matches the original label
122 if ((*host_cloud_)[i].label ==
123 (*host_cloud_)[data[qp_r + qp * max_answers]].label) {
124 processed[data[qp_r + qp * max_answers]] = true;
125 PointT t_l = (*host_cloud_)[data[qp_r + qp * max_answers]];
126 PointXYZ p_l;
127 p_l.x = t_l.x;
128 p_l.y = t_l.y;
129 p_l.z = t_l.z;
130 queries_host.push_back(p_l);
131 found_points++;
132 r.indices.push_back(data[qp_r + qp * max_answers]);
133 }
134 }
135 }
136 }
137 // If this queue is satisfactory, add to the clusters
138 if (found_points >= min_pts_per_cluster && found_points <= max_pts_per_cluster) {
139 std::sort(r.indices.begin(), r.indices.end());
140 // @todo: check if the following is actually still needed
141 // r.indices.erase (std::unique (r.indices.begin (), r.indices.end ()),
142 // r.indices.end ());
143
144 r.header = host_cloud_->header;
145 clusters.push_back(r); // We could avoid a copy by working directly in the vector
146 }
147 }
148}
149
150template <typename PointT>
151void
153 std::vector<PointIndices>& clusters)
154{
155 // Initialize the GPU search tree
156 if (!tree_) {
157 tree_.reset(new pcl::gpu::Octree());
158 ///@todo what do we do if input isn't a PointXYZ cloud?
159 tree_->setCloud(input_);
160 }
161 if (!tree_->isBuilt()) {
162 tree_->build();
163 }
164 /*
165 if(tree_->cloud_.size() != host_cloud.size ())
166 {
167 PCL_ERROR("[pcl::gpu::EuclideanClusterExtraction] size of host cloud and device
168 cloud don't match!\n"); return;
169 }
170 */
171 // Extract the actual clusters
172 extractLabeledEuclideanClusters<PointT>(host_cloud_,
173 tree_,
174 cluster_tolerance_,
175 clusters,
176 min_pts_per_cluster_,
177 max_pts_per_cluster_);
178
179 // Sort the clusters based on their size (largest one first)
180 std::sort(clusters.rbegin(), clusters.rend(), compareLabeledPointClusters);
181}
182
183#define PCL_INSTANTIATE_extractLabeledEuclideanClusters(T) \
184 template void PCL_EXPORTS pcl::gpu::extractLabeledEuclideanClusters<T>( \
185 const typename pcl::PointCloud<T>::Ptr&, \
186 const pcl::gpu::Octree::Ptr&, \
187 float, \
188 std::vector<PointIndices>&, \
189 unsigned int, \
190 unsigned int);
191#define PCL_INSTANTIATE_EuclideanLabeledClusterExtraction(T) \
192 template class PCL_EXPORTS pcl::gpu::EuclideanLabeledClusterExtraction<T>;
void push_back(const PointT &pt)
Insert a new point in the cloud, at the end of the container.
Definition: point_cloud.h:652
pcl::PCLHeader header
The point cloud header.
Definition: point_cloud.h:392
std::size_t size() const
Definition: point_cloud.h:443
std::vector< PointT, Eigen::aligned_allocator< PointT > > VectorType
Definition: point_cloud.h:411
shared_ptr< PointCloud< PointT > > Ptr
Definition: point_cloud.h:413
void upload(const T *host_ptr, std::size_t size)
Uploads data to internal buffer in GPU memory.
void download(T *host_ptr) const
Downloads data from internal buffer to CPU memory.
void extract(std::vector< PointIndices > &clusters)
extract clusters of a PointCloud given by <setInputCloud(), setIndices()>
Octree implementation on GPU.
Definition: octree.hpp:58
shared_ptr< Octree > Ptr
Types.
Definition: octree.hpp:68
bool compareLabeledPointClusters(const pcl::PointIndices &a, const pcl::PointIndices &b)
Sort clusters method (for std::sort).
void extractLabeledEuclideanClusters(const typename pcl::PointCloud< PointT >::Ptr &host_cloud_, const pcl::gpu::Octree::Ptr &tree, float tolerance, std::vector< PointIndices > &clusters, unsigned int min_pts_per_cluster, unsigned int max_pts_per_cluster)
::pcl::PCLHeader header
Definition: PointIndices.h:19
A point structure representing Euclidean xyz coordinates.
A point structure representing Euclidean xyz coordinates, and the RGB color.
DeviceArray< int > sizes
DeviceArray< int > data