Point Cloud Library (PCL) 1.13.0
Loading...
Searching...
No Matches
conversions.h
1/*
2 * Software License Agreement (BSD License)
3 *
4 * Point Cloud Library (PCL) - www.pointclouds.org
5 * Copyright (c) 2010-2012, Willow Garage, Inc.
6 *
7 * All rights reserved.
8 *
9 * Redistribution and use in source and binary forms, with or without
10 * modification, are permitted provided that the following conditions
11 * are met:
12 *
13 * * Redistributions of source code must retain the above copyright
14 * notice, this list of conditions and the following disclaimer.
15 * * Redistributions in binary form must reproduce the above
16 * copyright notice, this list of conditions and the following
17 * disclaimer in the documentation and/or other materials provided
18 * with the distribution.
19 * * Neither the name of the copyright holder(s) nor the names of its
20 * contributors may be used to endorse or promote products derived
21 * from this software without specific prior written permission.
22 *
23 * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS
24 * "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT
25 * LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS
26 * FOR A PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE
27 * COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT,
28 * INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING,
29 * BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
30 * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
31 * CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT
32 * LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN
33 * ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF ADVISED OF THE
34 * POSSIBILITY OF SUCH DAMAGE.
35 *
36 * $Id$
37 *
38 */
39
40#pragma once
41
42#ifdef __GNUC__
43#pragma GCC system_header
44#endif
45
46#include <pcl/PCLPointField.h>
47#include <pcl/PCLPointCloud2.h>
48#include <pcl/PCLImage.h>
49#include <pcl/point_cloud.h>
50#include <pcl/type_traits.h>
51#include <pcl/for_each_type.h>
52#include <pcl/console/print.h>
53
54#include <algorithm>
55#include <iterator>
56
57namespace pcl
58{
59 namespace detail
60 {
61 // For converting template point cloud to message.
62 template<typename PointT>
64 {
65 FieldAdder (std::vector<pcl::PCLPointField>& fields) : fields_ (fields) {};
66
67 template<typename U> void operator() ()
68 {
70 f.name = pcl::traits::name<PointT, U>::value;
71 f.offset = pcl::traits::offset<PointT, U>::value;
72 f.datatype = pcl::traits::datatype<PointT, U>::value;
73 f.count = pcl::traits::datatype<PointT, U>::size;
74 fields_.push_back (f);
75 }
76
77 std::vector<pcl::PCLPointField>& fields_;
78 };
79
80 // For converting message to template point cloud.
81 template<typename PointT>
83 {
84 FieldMapper (const std::vector<pcl::PCLPointField>& fields,
85 std::vector<FieldMapping>& map)
86 : fields_ (fields), map_ (map)
87 {
88 }
89
90 template<typename Tag> void
92 {
93 for (const auto& field : fields_)
94 {
95 if (FieldMatches<PointT, Tag>()(field))
96 {
97 FieldMapping mapping;
98 mapping.serialized_offset = field.offset;
99 mapping.struct_offset = pcl::traits::offset<PointT, Tag>::value;
100 mapping.size = sizeof (typename pcl::traits::datatype<PointT, Tag>::type);
101 map_.push_back (mapping);
102 return;
103 }
104 }
105 // Disable thrown exception per #595: http://dev.pointclouds.org/issues/595
106 PCL_WARN ("Failed to find match for field '%s'.\n", pcl::traits::name<PointT, Tag>::value);
107 //throw pcl::InvalidConversionException (ss.str ());
108 }
109
110 const std::vector<pcl::PCLPointField>& fields_;
111 std::vector<FieldMapping>& map_;
112 };
113
114 inline bool
116 {
118 }
119
120 } //namespace detail
121
122 template<typename PointT> void
123 createMapping (const std::vector<pcl::PCLPointField>& msg_fields, MsgFieldMap& field_map)
124 {
125 // Create initial 1-1 mapping between serialized data segments and struct fields
126 detail::FieldMapper<PointT> mapper (msg_fields, field_map);
127 for_each_type< typename traits::fieldList<PointT>::type > (mapper);
128
129 // Coalesce adjacent fields into single memcpy's where possible
130 if (field_map.size() > 1)
131 {
132 std::sort(field_map.begin(), field_map.end(), detail::fieldOrdering);
133 MsgFieldMap::iterator i = field_map.begin(), j = i + 1;
134 while (j != field_map.end())
135 {
136 // This check is designed to permit padding between adjacent fields.
137 /// @todo One could construct a pathological case where the struct has a
138 /// field where the serialized data has padding
139 if (j->serialized_offset - i->serialized_offset == j->struct_offset - i->struct_offset)
140 {
141 i->size += (j->struct_offset + j->size) - (i->struct_offset + i->size);
142 j = field_map.erase(j);
143 }
144 else
145 {
146 ++i;
147 ++j;
148 }
149 }
150 }
151 }
152
153 /** \brief Convert a PCLPointCloud2 binary data blob into a pcl::PointCloud<T> object using a field_map.
154 * \param[in] msg the PCLPointCloud2 binary blob
155 * \param[out] cloud the resultant pcl::PointCloud<T>
156 * \param[in] field_map a MsgFieldMap object
157 *
158 * \note Use fromPCLPointCloud2 (PCLPointCloud2, PointCloud<T>) directly or create you
159 * own MsgFieldMap using:
160 *
161 * \code
162 * MsgFieldMap field_map;
163 * createMapping<PointT> (msg.fields, field_map);
164 * \endcode
165 */
166 template <typename PointT> void
168 const MsgFieldMap& field_map)
169 {
170 // Copy info fields
171 cloud.header = msg.header;
172 cloud.width = msg.width;
173 cloud.height = msg.height;
174 cloud.is_dense = msg.is_dense == 1;
175
176 // Copy point data
177 cloud.resize (msg.width * msg.height);
178 std::uint8_t* cloud_data = reinterpret_cast<std::uint8_t*>(&cloud[0]);
179
180 // Check if we can copy adjacent points in a single memcpy. We can do so if there
181 // is exactly one field to copy and it is the same size as the source and destination
182 // point types.
183 if (field_map.size() == 1 &&
184 field_map[0].serialized_offset == 0 &&
185 field_map[0].struct_offset == 0 &&
186 field_map[0].size == msg.point_step &&
187 field_map[0].size == sizeof(PointT))
188 {
189 const auto cloud_row_step = (sizeof (PointT) * cloud.width);
190 const std::uint8_t* msg_data = &msg.data[0];
191 // Should usually be able to copy all rows at once
192 if (msg.row_step == cloud_row_step)
193 {
194 std::copy(msg.data.cbegin(), msg.data.cend(), cloud_data);
195 }
196 else
197 {
198 for (uindex_t i = 0; i < msg.height; ++i, cloud_data += cloud_row_step, msg_data += msg.row_step)
199 memcpy (cloud_data, msg_data, cloud_row_step);
200 }
201
202 }
203 else
204 {
205 // If not, memcpy each group of contiguous fields separately
206 for (uindex_t row = 0; row < msg.height; ++row)
207 {
208 const std::uint8_t* row_data = &msg.data[row * msg.row_step];
209 for (uindex_t col = 0; col < msg.width; ++col)
210 {
211 const std::uint8_t* msg_data = row_data + col * msg.point_step;
212 for (const detail::FieldMapping& mapping : field_map)
213 {
214 std::copy(msg_data + mapping.serialized_offset, msg_data + mapping.serialized_offset + mapping.size,
215 cloud_data + mapping.struct_offset);
216 }
217 cloud_data += sizeof (PointT);
218 }
219 }
220 }
221 }
222
223 /** \brief Convert a PCLPointCloud2 binary data blob into a pcl::PointCloud<T> object.
224 * \param[in] msg the PCLPointCloud2 binary blob
225 * \param[out] cloud the resultant pcl::PointCloud<T>
226 */
227 template<typename PointT> void
229 {
230 MsgFieldMap field_map;
231 createMapping<PointT> (msg.fields, field_map);
232 fromPCLPointCloud2 (msg, cloud, field_map);
233 }
234
235 /** \brief Convert a pcl::PointCloud<T> object to a PCLPointCloud2 binary data blob.
236 * \param[in] cloud the input pcl::PointCloud<T>
237 * \param[out] msg the resultant PCLPointCloud2 binary blob
238 */
239 template<typename PointT> void
241 {
242 // Ease the user's burden on specifying width/height for unorganized datasets
243 if (cloud.width == 0 && cloud.height == 0)
244 {
245 msg.width = cloud.size ();
246 msg.height = 1;
247 }
248 else
249 {
250 assert (cloud.size () == cloud.width * cloud.height);
251 msg.height = cloud.height;
252 msg.width = cloud.width;
253 }
254
255 // Fill point cloud binary data (padding and all)
256 std::size_t data_size = sizeof (PointT) * cloud.size ();
257 msg.data.resize (data_size);
258 if (data_size)
259 {
260 memcpy(&msg.data[0], &cloud[0], data_size);
261 }
262
263 // Fill fields metadata
264 msg.fields.clear ();
265 for_each_type<typename traits::fieldList<PointT>::type> (detail::FieldAdder<PointT>(msg.fields));
266
267 msg.header = cloud.header;
268 msg.point_step = sizeof (PointT);
269 msg.row_step = (sizeof (PointT) * msg.width);
270 msg.is_dense = cloud.is_dense;
271 /// @todo msg.is_bigendian = ?;
272 }
273
274 /** \brief Copy the RGB fields of a PointCloud into pcl::PCLImage format
275 * \param[in] cloud the point cloud message
276 * \param[out] msg the resultant pcl::PCLImage
277 * CloudT cloud type, CloudT should be akin to pcl::PointCloud<pcl::PointXYZRGBA>
278 * \note will throw std::runtime_error if there is a problem
279 */
280 template<typename CloudT> void
281 toPCLPointCloud2 (const CloudT& cloud, pcl::PCLImage& msg)
282 {
283 // Ease the user's burden on specifying width/height for unorganized datasets
284 if (cloud.width == 0 && cloud.height == 0)
285 throw std::runtime_error("Needs to be a dense like cloud!!");
286 else
287 {
288 if (cloud.size () != cloud.width * cloud.height)
289 throw std::runtime_error("The width and height do not match the cloud size!");
290 msg.height = cloud.height;
291 msg.width = cloud.width;
292 }
293
294 // ensor_msgs::image_encodings::BGR8;
295 msg.header = cloud.header;
296 msg.encoding = "bgr8";
297 msg.step = msg.width * sizeof (std::uint8_t) * 3;
298 msg.data.resize (msg.step * msg.height);
299 for (std::size_t y = 0; y < cloud.height; y++)
300 {
301 for (std::size_t x = 0; x < cloud.width; x++)
302 {
303 std::uint8_t * pixel = &(msg.data[y * msg.step + x * 3]);
304 memcpy (pixel, &cloud (x, y).rgb, 3 * sizeof(std::uint8_t));
305 }
306 }
307 }
308
309 /** \brief Copy the RGB fields of a PCLPointCloud2 msg into pcl::PCLImage format
310 * \param cloud the point cloud message
311 * \param msg the resultant pcl::PCLImage
312 * will throw std::runtime_error if there is a problem
313 */
314 inline void
316 {
317 const auto predicate = [](const auto& field) { return field.name == "rgb"; };
318 const auto result = std::find_if(cloud.fields.cbegin (), cloud.fields.cend (), predicate);
319 if (result == cloud.fields.end ())
320 throw std::runtime_error ("No rgb field!!");
321
322 const auto rgb_index = std::distance(cloud.fields.begin (), result);
323 if (cloud.width == 0 && cloud.height == 0)
324 throw std::runtime_error ("Needs to be a dense like cloud!!");
325 else
326 {
327 msg.height = cloud.height;
328 msg.width = cloud.width;
329 }
330 auto rgb_offset = cloud.fields[rgb_index].offset;
331 const auto point_step = cloud.point_step;
332
333 // pcl::image_encodings::BGR8;
334 msg.header = cloud.header;
335 msg.encoding = "bgr8";
336 msg.step = (msg.width * sizeof (std::uint8_t) * 3);
337 msg.data.resize (msg.step * msg.height);
338
339 for (std::size_t y = 0; y < cloud.height; y++)
340 {
341 for (std::size_t x = 0; x < cloud.width; x++, rgb_offset += point_step)
342 {
343 std::uint8_t * pixel = &(msg.data[y * msg.step + x * 3]);
344 std::copy(&cloud.data[rgb_offset], &cloud.data[rgb_offset] + 3, pixel);
345 }
346 }
347 }
348}
PointCloud represents the base class in PCL for storing collections of 3D points.
bool is_dense
True if no points are invalid (e.g., have NaN or Inf values in any of their floating point fields).
void resize(std::size_t count)
Resizes the container to contain count elements.
std::uint32_t width
The point cloud width (if organized as an image-structure).
pcl::PCLHeader header
The point cloud header.
std::uint32_t height
The point cloud height (if organized as an image-structure).
std::size_t size() const
bool fieldOrdering(const FieldMapping &a, const FieldMapping &b)
detail::int_type_t< detail::index_type_size, false > uindex_t
Type used for an unsigned index in PCL.
Definition types.h:120
void toPCLPointCloud2(const pcl::PointCloud< PointT > &cloud, pcl::PCLPointCloud2 &msg)
Convert a pcl::PointCloud<T> object to a PCLPointCloud2 binary data blob.
void createMapping(const std::vector< pcl::PCLPointField > &msg_fields, MsgFieldMap &field_map)
void fromPCLPointCloud2(const pcl::PCLPointCloud2 &msg, pcl::PointCloud< PointT > &cloud, const MsgFieldMap &field_map)
Convert a PCLPointCloud2 binary data blob into a pcl::PointCloud<T> object using a field_map.
std::vector< detail::FieldMapping > MsgFieldMap
Definition point_cloud.h:72
uindex_t step
Definition PCLImage.h:21
uindex_t height
Definition PCLImage.h:16
std::string encoding
Definition PCLImage.h:18
std::vector< std::uint8_t > data
Definition PCLImage.h:23
uindex_t width
Definition PCLImage.h:17
::pcl::PCLHeader header
Definition PCLImage.h:14
std::uint8_t is_dense
std::vector<::pcl::PCLPointField > fields
::pcl::PCLHeader header
std::vector< std::uint8_t > data
std::uint8_t datatype
A point structure representing Euclidean xyz coordinates, and the RGB color.
FieldAdder(std::vector< pcl::PCLPointField > &fields)
Definition conversions.h:65
std::vector< pcl::PCLPointField > & fields_
Definition conversions.h:77
FieldMapper(const std::vector< pcl::PCLPointField > &fields, std::vector< FieldMapping > &map)
Definition conversions.h:84
const std::vector< pcl::PCLPointField > & fields_
std::vector< FieldMapping > & map_
std::size_t serialized_offset
Definition point_cloud.h:64