Point Cloud Library (PCL) 1.12.1
point_cloud_image_extractors.hpp
1/*
2 * Software License Agreement (BSD License)
3 *
4 * Point Cloud Library (PCL) - www.pointclouds.org
5 * Copyright (c) 2013-, Open Perception, 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 */
37
38#pragma once
39
40#include <set>
41#include <map>
42#include <ctime>
43#include <cstdlib>
44
45#include <pcl/common/io.h>
46#include <pcl/common/colors.h>
47#include <pcl/common/point_tests.h> // for pcl::isFinite
48
49///////////////////////////////////////////////////////////////////////////////////////////
50template <typename PointT> bool
52{
53 if (!cloud.isOrganized () || cloud.size () != cloud.width * cloud.height)
54 return (false);
55
56 bool result = this->extractImpl (cloud, img);
57
58 if (paint_nans_with_black_ && result)
59 {
60 std::size_t size = img.encoding == "mono16" ? 2 : 3;
61 for (std::size_t i = 0; i < cloud.size (); ++i)
62 if (!pcl::isFinite (cloud[i]))
63 std::memset (&img.data[i * size], 0, size);
64 }
65
66 return (result);
67}
68
69///////////////////////////////////////////////////////////////////////////////////////////
70template <typename PointT> bool
72{
73 std::vector<pcl::PCLPointField> fields;
74 int field_x_idx = pcl::getFieldIndex<PointT> ("normal_x", fields);
75 int field_y_idx = pcl::getFieldIndex<PointT> ("normal_y", fields);
76 int field_z_idx = pcl::getFieldIndex<PointT> ("normal_z", fields);
77 if (field_x_idx == -1 || field_y_idx == -1 || field_z_idx == -1)
78 return (false);
79 const std::size_t offset_x = fields[field_x_idx].offset;
80 const std::size_t offset_y = fields[field_y_idx].offset;
81 const std::size_t offset_z = fields[field_z_idx].offset;
82
83 img.encoding = "rgb8";
84 img.width = cloud.width;
85 img.height = cloud.height;
86 img.step = img.width * sizeof (unsigned char) * 3;
87 img.data.resize (img.step * img.height);
88
89 for (std::size_t i = 0; i < cloud.size (); ++i)
90 {
91 float x;
92 float y;
93 float z;
94 pcl::getFieldValue<PointT, float> (cloud[i], offset_x, x);
95 pcl::getFieldValue<PointT, float> (cloud[i], offset_y, y);
96 pcl::getFieldValue<PointT, float> (cloud[i], offset_z, z);
97 img.data[i * 3 + 0] = static_cast<unsigned char>((x + 1.0) * 127);
98 img.data[i * 3 + 1] = static_cast<unsigned char>((y + 1.0) * 127);
99 img.data[i * 3 + 2] = static_cast<unsigned char>((z + 1.0) * 127);
100 }
101
102 return (true);
103}
104
105///////////////////////////////////////////////////////////////////////////////////////////
106template <typename PointT> bool
108{
109 std::vector<pcl::PCLPointField> fields;
110 int field_idx = pcl::getFieldIndex<PointT> ("rgb", fields);
111 if (field_idx == -1)
112 {
113 field_idx = pcl::getFieldIndex<PointT> ("rgba", fields);
114 if (field_idx == -1)
115 return (false);
116 }
117 const std::size_t offset = fields[field_idx].offset;
118
119 img.encoding = "rgb8";
120 img.width = cloud.width;
121 img.height = cloud.height;
122 img.step = img.width * sizeof (unsigned char) * 3;
123 img.data.resize (img.step * img.height);
124
125 for (std::size_t i = 0; i < cloud.size (); ++i)
126 {
127 std::uint32_t val;
128 pcl::getFieldValue<PointT, std::uint32_t> (cloud[i], offset, val);
129 img.data[i * 3 + 0] = (val >> 16) & 0x0000ff;
130 img.data[i * 3 + 1] = (val >> 8) & 0x0000ff;
131 img.data[i * 3 + 2] = (val) & 0x0000ff;
132 }
133
134 return (true);
135}
136
137///////////////////////////////////////////////////////////////////////////////////////////
138template <typename PointT> bool
140{
141 std::vector<pcl::PCLPointField> fields;
142 int field_idx = pcl::getFieldIndex<PointT> ("label", fields);
143 if (field_idx == -1)
144 return (false);
145 const std::size_t offset = fields[field_idx].offset;
146
147 switch (color_mode_)
148 {
149 case COLORS_MONO:
150 {
151 img.encoding = "mono16";
152 img.width = cloud.width;
153 img.height = cloud.height;
154 img.step = img.width * sizeof (unsigned short);
155 img.data.resize (img.step * img.height);
156 unsigned short* data = reinterpret_cast<unsigned short*>(&img.data[0]);
157 for (std::size_t i = 0; i < cloud.size (); ++i)
158 {
159 std::uint32_t val;
160 pcl::getFieldValue<PointT, std::uint32_t> (cloud[i], offset, val);
161 data[i] = static_cast<unsigned short>(val);
162 }
163 break;
164 }
165 case COLORS_RGB_RANDOM:
166 {
167 img.encoding = "rgb8";
168 img.width = cloud.width;
169 img.height = cloud.height;
170 img.step = img.width * sizeof (unsigned char) * 3;
171 img.data.resize (img.step * img.height);
172
173 std::srand(std::time(nullptr));
174 std::map<std::uint32_t, std::size_t> colormap;
175
176 for (std::size_t i = 0; i < cloud.size (); ++i)
177 {
178 std::uint32_t val;
179 pcl::getFieldValue<PointT, std::uint32_t> (cloud[i], offset, val);
180 if (colormap.count (val) == 0)
181 {
182 colormap[val] = i * 3;
183 img.data[i * 3 + 0] = static_cast<std::uint8_t> ((std::rand () % 256));
184 img.data[i * 3 + 1] = static_cast<std::uint8_t> ((std::rand () % 256));
185 img.data[i * 3 + 2] = static_cast<std::uint8_t> ((std::rand () % 256));
186 }
187 else
188 {
189 memcpy (&img.data[i * 3], &img.data[colormap[val]], 3);
190 }
191 }
192 break;
193 }
194 case COLORS_RGB_GLASBEY:
195 {
196 img.encoding = "rgb8";
197 img.width = cloud.width;
198 img.height = cloud.height;
199 img.step = img.width * sizeof (unsigned char) * 3;
200 img.data.resize (img.step * img.height);
201
202 std::srand(std::time(nullptr));
203 std::set<std::uint32_t> labels;
204 std::map<std::uint32_t, std::size_t> colormap;
205
206 // First pass: find unique labels
207 for (const auto& point: cloud)
208 {
209 // If we need to paint NaN points with black do not waste colors on them
210 if (paint_nans_with_black_ && !pcl::isFinite (point))
211 continue;
212 std::uint32_t val;
213 pcl::getFieldValue<PointT, std::uint32_t> (point, offset, val);
214 labels.insert (val);
215 }
216
217 // Assign Glasbey colors in ascending order of labels
218 // Note: the color LUT has a finite size (256 colors), therefore when
219 // there are more labels the colors will repeat
220 std::size_t color = 0;
221 for (const std::uint32_t &label : labels)
222 {
223 colormap[label] = color % GlasbeyLUT::size ();
224 ++color;
225 }
226
227 // Second pass: copy colors from the LUT
228 for (std::size_t i = 0; i < cloud.size (); ++i)
229 {
230 std::uint32_t val;
231 pcl::getFieldValue<PointT, std::uint32_t> (cloud[i], offset, val);
232 memcpy (&img.data[i * 3], GlasbeyLUT::data () + colormap[val] * 3, 3);
233 }
234
235 break;
236 }
237 }
238
239 return (true);
240}
241
242///////////////////////////////////////////////////////////////////////////////////////////
243template <typename PointT> bool
245{
246 std::vector<pcl::PCLPointField> fields;
247 int field_idx = pcl::getFieldIndex<PointT> (field_name_, fields);
248 if (field_idx == -1)
249 return (false);
250 const std::size_t offset = fields[field_idx].offset;
251
252 img.encoding = "mono16";
253 img.width = cloud.width;
254 img.height = cloud.height;
255 img.step = img.width * sizeof (unsigned short);
256 img.data.resize (img.step * img.height);
257 unsigned short* data = reinterpret_cast<unsigned short*>(&img.data[0]);
258
259 float scaling_factor = scaling_factor_;
260 float data_min = 0.0f;
261 if (scaling_method_ == SCALING_FULL_RANGE)
262 {
263 float min = std::numeric_limits<float>::infinity();
264 float max = -std::numeric_limits<float>::infinity();
265 for (const auto& point: cloud)
266 {
267 float val;
268 pcl::getFieldValue<PointT, float> (point, offset, val);
269 if (val < min)
270 min = val;
271 if (val > max)
272 max = val;
273 }
274 scaling_factor = min == max ? 0 : std::numeric_limits<unsigned short>::max() / (max - min);
275 data_min = min;
276 }
277
278 for (std::size_t i = 0; i < cloud.size (); ++i)
279 {
280 float val;
281 pcl::getFieldValue<PointT, float> (cloud[i], offset, val);
282 if (scaling_method_ == SCALING_NO)
283 {
284 data[i] = val;
285 }
286 else if (scaling_method_ == SCALING_FULL_RANGE)
287 {
288 data[i] = (val - data_min) * scaling_factor;
289 }
290 else if (scaling_method_ == SCALING_FIXED_FACTOR)
291 {
292 data[i] = val * scaling_factor;
293 }
294 }
295
296 return (true);
297}
298
PointCloud represents the base class in PCL for storing collections of 3D points.
Definition: point_cloud.h:173
std::uint32_t width
The point cloud width (if organized as an image-structure).
Definition: point_cloud.h:398
bool isOrganized() const
Return whether a dataset is organized (e.g., arranged in a structured grid).
Definition: point_cloud.h:310
std::uint32_t height
The point cloud height (if organized as an image-structure).
Definition: point_cloud.h:400
std::size_t size() const
Definition: point_cloud.h:443
bool extractImpl(const PointCloud &cloud, pcl::PCLImage &img) const override
Implementation of the extract() function, has to be implemented in deriving classes.
bool extractImpl(const PointCloud &cloud, pcl::PCLImage &img) const override
Implementation of the extract() function, has to be implemented in deriving classes.
bool extractImpl(const PointCloud &cloud, pcl::PCLImage &img) const override
Implementation of the extract() function, has to be implemented in deriving classes.
bool extract(const PointCloud &cloud, pcl::PCLImage &image) const
Obtain the image from the given cloud.
bool extractImpl(const PointCloud &cloud, pcl::PCLImage &image) const override
Implementation of the extract() function, has to be implemented in deriving classes.
bool isFinite(const PointT &pt)
Tests if the 3D components of a point are all finite param[in] pt point to be tested return true if f...
Definition: point_tests.h:55
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