Installing Point Cloud Library

Recently I had to help someone get the Point Cloud Library and sample applications working under Linux using the Open NI Framework and a Microsoft Kinect for the XBOX 360. This isn’t exactly something you might want to do every day but the results were quite ‘interesting’ and besides I got it working…

As anyone who hasn’t been living under a rock for the last five years knows the Kinect is Microsoft’s revolutionary ‘3D’ motion sensing controller – what is not as well-known is that the Point Cloud library provides a set of Open Source routines for 2D and 3D image processing. The aim was to use the Open NI Framework to allow a Kinect to be used to capture a 3D ‘image’ of it’s surroundings.

Note – The instructions below assume you are using Linux Mint 17, in theory they should apply to Ubuntu and Debian, but I was unable to install the Point Cloud Library on Debian ‘wheezy’.

Connecting the Kinect

The Kinect I had was supplied with an adapter cable and power supply to allow it to be connected to a PC, but since there is more than one type of Kinect and the installation process involves downloading and installing the correct version of the sensor module, it is worth checking that you have the same type of Kinect and that it is connected to your PC and is being detected – to do this plug it in, wait a few seconds then list all the USB devices. This can save a lot of problems later!

$ lsusb
Bus 002 Device 002: ID 8087:0024 Intel Corp. Integrated Rate Matching Hub
Bus 002 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 001 Device 004: ID 1bcf:2c08 Sunplus Innovation Technology Inc.
Bus 001 Device 003: ID 08ff:168f AuthenTec, Inc. AES1660 Fingerprint Sensor
Bus 001 Device 002: ID 8087:0024 Intel Corp. Integrated Rate Matching Hub
Bus 001 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub
Bus 004 Device 001: ID 1d6b:0003 Linux Foundation 3.0 root hub
Bus 003 Device 006: ID 045e:02bf Microsoft Corp.
Bus 003 Device 005: ID 045e:02be Microsoft Corp.
Bus 003 Device 004: ID 045e:02c2 Microsoft Corp.
Bus 003 Device 003: ID 0951:1689 Kingston Technology DataTraveler SE9
Bus 003 Device 002: ID 046d:c018 Logitech, Inc. Optical Wheel Mouse
Bus 003 Device 001: ID 1d6b:0002 Linux Foundation 2.0 root hub

The IDs shown in italics should match those of your Kinect, if you cannot see any devices check that your Kinect is connected properly and powered on. If the IDs for the Kinect are different then you have a different model and will need a different version of the sensor module.

Installing the required tools

There are some additional tools and libraries we will need to download the sensor module and compile the Point Cloud Library sample code later.

# apt-get install cmake
# apt-get install g++
# apt-get install libboost-all-dev
# apt-get install git


Installing OpenNI

Finally we are ready to start installing the Point Cloud Library and Open NI Framework, fortunately on current versions of Linux most of the software we need is available as prebuilt packages so we just need to install them – but this needs to be done in the right order, beginning with the Open NI Framework.

# apt-get install libopenni-dev
# apt-get install libopenni0
# apt-get install openni-utils

Note: It looks like these modules are built with SSE3 support enabled, if you get an ‘Illegal instruction’ error at runtime you will need to edit the make files and recompile them from the source without SSE3 support.

Then we need to install the modified sensor module from GitHub as it isn’t available in the repositories.

# git clone https://github.com/avin2/SensorKinect

Obviously if the version has been updated since I wrote this post the exact the names of the files below may not quite be the same, and need to make sure you select the right architecture for your machine (x86 or x64).

For x64 systems:

# bzip2 -d ./SensorKinect/Bin/SensorKinect093-Bin-Linux-x64-v5.1.2.1.tar.bz2
# tar -xvf ./SensorKinect/Bin/SensorKinect093-Bin-Linux-x64-v5.1.2.1.tar
# cd Sensor-Bin-Linux-x64-v5.1.2.1
# ./install.sh

For x86 systems:

# bzip2 -d ./SensorKinect/Bin/SensorKinect093-Bin-Linux-x86-v5.1.2.1.tar.bz2
# tar -xvf ./SensorKinect/Bin/SensorKinect093-Bin-Linux-x86-v5.1.2.1.tar
# cd Sensor-Bin-Linux-x86-v5.1.2.1
# ./install.sh


Installing the Point Cloud Library

Like the sensor module the Point Cloud Libraries are not available in the default repositories, but fortunately the necessary packages can be installed using these instructions. As I was using Linux Mint 17 I followed the instructions for Ubuntu to add the appropriate PPA to the apt repositories.

# add-apt-repository ppa:v-launchpad-jochen-sprickerhof-de/pcl
# apt-get update

Then all I had to do was to install the Point Cloud Library…

# apt-get install libpcl-all
  :
  :
  :

The installation will take a while so now would be a good time to grab a coffee…


Testing

If that all worked then you can try using the ‘pcl_viewer’ to view an example dataset

$ vi pcd_example.pcd

# .PCD v.7 - Point Cloud Data file format
VERSION .7
FIELDS x y z rgb
SIZE 4 4 4 4
TYPE F F F F
COUNT 1 1 1 1
WIDTH 213
HEIGHT 1
VIEWPOINT 0 0 0 1 0 0 0
POINTS 213
DATA ascii
0.93773 0.33763 0 4.2108e+06
0.90805 0.35641 0 4.2108e+06
0.81915 0.32 0 4.2108e+06
0.97192 0.278 0 4.2108e+06
0.944 0.29474 0 4.2108e+06
0.98111 0.24247 0 4.2108e+06
0.93655 0.26143 0 4.2108e+06
0.91631 0.27442 0 4.2108e+06
0.81921 0.29315 0 4.2108e+06
0.90701 0.24109 0 4.2108e+06
0.83239 0.23398 0 4.2108e+06
0.99185 0.2116 0 4.2108e+06
0.89264 0.21174 0 4.2108e+06
0.85082 0.21212 0 4.2108e+06
0.81044 0.32222 0 4.2108e+06
0.74459 0.32192 0 4.2108e+06
0.69927 0.32278 0 4.2108e+06
0.8102 0.29315 0 4.2108e+06
0.75504 0.29765 0 4.2108e+06
0.8102 0.24399 0 4.2108e+06
0.74995 0.24723 0 4.2108e+06
0.68049 0.29768 0 4.2108e+06
0.66509 0.29002 0 4.2108e+06
0.69441 0.2526 0 4.2108e+06
0.62807 0.22187 0 4.2108e+06
0.58706 0.32199 0 4.2108e+06
0.52125 0.31955 0 4.2108e+06
0.49351 0.32282 0 4.2108e+06
0.44313 0.32169 0 4.2108e+06
0.58678 0.2929 0 4.2108e+06
0.53436 0.29164 0 4.2108e+06
0.59308 0.24134 0 4.2108e+06
0.5357 0.2444 0 4.2108e+06
0.50043 0.31235 0 4.2108e+06
0.44107 0.29711 0 4.2108e+06
0.50727 0.22193 0 4.2108e+06
0.43957 0.23976 0 4.2108e+06
0.8105 0.21112 0 4.2108e+06
0.73555 0.2114 0 4.2108e+06
0.69907 0.21082 0 4.2108e+06
0.63327 0.21154 0 4.2108e+06
0.59165 0.21201 0 4.2108e+06
0.52477 0.21491 0 4.2108e+06
0.49375 0.21006 0 4.2108e+06
0.4384 0.19632 0 4.2108e+06
0.43425 0.16052 0 4.2108e+06
0.3787 0.32173 0 4.2108e+06
0.33444 0.3216 0 4.2108e+06
0.23815 0.32199 0 4.808e+06
0.3788 0.29315 0 4.2108e+06
0.33058 0.31073 0 4.2108e+06
0.3788 0.24399 0 4.2108e+06
0.30249 0.29189 0 4.2108e+06
0.23492 0.29446 0 4.808e+06
0.29465 0.24399 0 4.2108e+06
0.23514 0.24172 0 4.808e+06
0.18836 0.32277 0 4.808e+06
0.15992 0.32176 0 4.808e+06
0.08642 0.32181 0 4.808e+06
0.039994 0.32283 0 4.808e+06
0.20039 0.31211 0 4.808e+06
0.1417 0.29506 0 4.808e+06
0.20921 0.22332 0 4.808e+06
0.13884 0.24227 0 4.808e+06
0.085123 0.29441 0 4.808e+06
0.048446 0.31279 0 4.808e+06
0.086957 0.24399 0 4.808e+06
0.3788 0.21189 0 4.2108e+06
0.29465 0.19323 0 4.2108e+06
0.23755 0.19348 0 4.808e+06
0.29463 0.16054 0 4.2108e+06
0.23776 0.16054 0 4.808e+06
0.19016 0.21038 0 4.808e+06
0.15704 0.21245 0 4.808e+06
0.08678 0.21169 0 4.808e+06
0.012746 0.32168 0 4.808e+06
-0.075715 0.32095 0 4.808e+06
-0.10622 0.32304 0 4.808e+06
-0.16391 0.32118 0 4.808e+06
0.00088411 0.29487 0 4.808e+06
-0.057568 0.29457 0 4.808e+06
-0.0034333 0.24399 0 4.808e+06
-0.055185 0.24185 0 4.808e+06
-0.10983 0.31352 0 4.808e+06
-0.15082 0.29453 0 4.808e+06
-0.11534 0.22049 0 4.808e+06
-0.15155 0.24381 0 4.808e+06
-0.1912 0.32173 0 4.808e+06
-0.281 0.3185 0 4.808e+06
-0.30791 0.32307 0 4.808e+06
-0.33854 0.32148 0 4.808e+06
-0.21248 0.29805 0 4.808e+06
-0.26372 0.29905 0 4.808e+06
-0.22562 0.24399 0 4.808e+06
-0.25035 0.2371 0 4.808e+06
-0.29941 0.31191 0 4.808e+06
-0.35845 0.2954 0 4.808e+06
-0.29231 0.22236 0 4.808e+06
-0.36101 0.24172 0 4.808e+06
-0.0034393 0.21129 0 4.808e+06
-0.07306 0.21304 0 4.808e+06
-0.10579 0.2099 0 4.808e+06
-0.13642 0.21411 0 4.808e+06
-0.22562 0.19323 0 4.808e+06
-0.24439 0.19799 0 4.808e+06
-0.22591 0.16041 0 4.808e+06
-0.23466 0.16082 0 4.808e+06
-0.3077 0.20998 0 4.808e+06
-0.3413 0.21239 0 4.808e+06
-0.40551 0.32178 0 4.2108e+06
-0.50568 0.3218 0 4.2108e+06
-0.41732 0.30844 0 4.2108e+06
-0.44237 0.28859 0 4.2108e+06
-0.41591 0.22004 0 4.2108e+06
-0.44803 0.24236 0 4.2108e+06
-0.50623 0.29315 0 4.2108e+06
-0.50916 0.24296 0 4.2108e+06
-0.57019 0.22334 0 4.2108e+06
-0.59611 0.32199 0 4.2108e+06
-0.65104 0.32199 0 4.2108e+06
-0.72566 0.32129 0 4.2108e+06
-0.75538 0.32301 0 4.2108e+06
-0.59653 0.29315 0 4.2108e+06
-0.65063 0.29315 0 4.2108e+06
-0.59478 0.24245 0 4.2108e+06
-0.65063 0.24399 0 4.2108e+06
-0.70618 0.29525 0 4.2108e+06
-0.76203 0.31284 0 4.2108e+06
-0.70302 0.24183 0 4.2108e+06
-0.77062 0.22133 0 4.2108e+06
-0.41545 0.21099 0 4.2108e+06
-0.45004 0.19812 0 4.2108e+06
-0.4475 0.1673 0 4.2108e+06
-0.52031 0.21236 0 4.2108e+06
-0.55182 0.21045 0 4.2108e+06
-0.5965 0.21131 0 4.2108e+06
-0.65064 0.2113 0 4.2108e+06
-0.72216 0.21286 0 4.2108e+06
-0.7556 0.20987 0 4.2108e+06
-0.78343 0.31973 0 4.2108e+06
-0.87572 0.32111 0 4.2108e+06
-0.90519 0.32263 0 4.2108e+06
-0.95526 0.34127 0 4.2108e+06
-0.79774 0.29271 0 4.2108e+06
-0.85618 0.29497 0 4.2108e+06
-0.79975 0.24326 0 4.2108e+06
-0.8521 0.24246 0 4.2108e+06
-0.91157 0.31224 0 4.2108e+06
-0.95031 0.29572 0 4.2108e+06
-0.92223 0.2213 0 4.2108e+06
-0.94979 0.24354 0 4.2108e+06
-0.78641 0.21505 0 4.2108e+06
-0.87094 0.21237 0 4.2108e+06
-0.90637 0.20934 0 4.2108e+06
-0.93777 0.21481 0 4.2108e+06
0.22244 -0.0296 0 4.808e+06
0.2704 -0.078167 0 4.808e+06
0.24416 -0.056883 0 4.808e+06
0.27311 -0.10653 0 4.808e+06
0.26172 -0.10653 0 4.808e+06
0.2704 -0.1349 0 4.808e+06
0.24428 -0.15599 0 4.808e+06
0.19017 -0.025297 0 4.808e+06
0.14248 -0.02428 0 4.808e+06
0.19815 -0.037432 0 4.808e+06
0.14248 -0.03515 0 4.808e+06
0.093313 -0.02428 0 4.808e+06
0.044144 -0.02428 0 4.808e+06
0.093313 -0.03515 0 4.808e+06
0.044144 -0.03515 0 4.808e+06
0.21156 -0.17357 0 4.808e+06
0.029114 -0.12594 0 4.2108e+06
0.036583 -0.15619 0 4.2108e+06
0.22446 -0.20514 0 4.808e+06
0.2208 -0.2369 0 4.808e+06
0.2129 -0.208 0 4.808e+06
0.19316 -0.25672 0 4.808e+06
0.14497 -0.27484 0 4.808e+06
0.030167 -0.18748 0 4.2108e+06
0.1021 -0.27453 0 4.808e+06
0.1689 -0.2831 0 4.808e+06
0.13875 -0.28647 0 4.808e+06
0.086993 -0.29568 0 4.808e+06
0.044924 -0.3154 0 4.808e+06
-0.0066125 -0.02428 0 4.808e+06
-0.057362 -0.02428 0 4.808e+06
-0.0066125 -0.03515 0 4.808e+06
-0.057362 -0.03515 0 4.808e+06
-0.10653 -0.02428 0 4.808e+06
-0.15266 -0.025282 0 4.808e+06
-0.10653 -0.03515 0 4.808e+06
-0.16036 -0.037257 0 4.808e+06
0.0083286 -0.1259 0 4.2108e+06
0.0007442 -0.15603 0 4.2108e+06
-0.1741 -0.17381 0 4.808e+06
-0.18502 -0.02954 0 4.808e+06
-0.20707 -0.056403 0 4.808e+06
-0.23348 -0.07764 0 4.808e+06
-0.2244 -0.10653 0 4.808e+06
-0.23604 -0.10652 0 4.808e+06
-0.20734 -0.15641 0 4.808e+06
-0.23348 -0.13542 0 4.808e+06
0.0061083 -0.18729 0 4.2108e+06
-0.066235 -0.27472 0 4.808e+06
-0.17577 -0.20789 0 4.808e+06
-0.10861 -0.27494 0 4.808e+06
-0.15584 -0.25716 0 4.808e+06
-0.0075775 -0.31546 0 4.808e+06
-0.050817 -0.29595 0 4.808e+06
-0.10306 -0.28653 0 4.808e+06
-0.1319 -0.2831 0 4.808e+06
-0.18716 -0.20571 0 4.808e+06
-0.18369 -0.23729 0 4.808e+06

$ pcl_viewer pcd_example.pcd

pcl-exampleYou may need to click on the image above to see the image properly. The display is reversed so you will need to rotate it to be able to read it easliy...

Compiling you own code

To compile your own code (in this case one of the Point Cloud Library tutorial programs) create a directory structure to hold the project.

$ cd ~
$ mkdir Code
$ cd Code
$ mkdir pcd_write
$ cd pcd_write

Then create a new file and in this case paste in the sample code.

$ vi pcd_write.cpp

#include <iostream>
#include <pcl/io/pcd_io.h>
#include <pcl/point_types.h>

int
  main (int argc, char** argv)
{
  pcl::PointCloud <pcl::PointXYZ> cloud;

  // Fill in the cloud data
  cloud.width    = 5;
  cloud.height   = 1;
  cloud.is_dense = false;
  cloud.points.resize (cloud.width * cloud.height);

  for (size_t i = 0; i < cloud.points.size (); ++i)
  {
    cloud.points[i].x = 1024 * rand () / (RAND_MAX + 1.0f);
    cloud.points[i].y = 1024 * rand () / (RAND_MAX + 1.0f);
    cloud.points[i].z = 1024 * rand () / (RAND_MAX + 1.0f);
  }

  pcl::io::savePCDFileASCII ("test_pcd.pcd", cloud);
  std::cerr << "Saved " << cloud.points.size () << " data points to test_pcd.pcd." << std::endl;

  for (size_t i = 0; i < cloud.points.size (); ++i)
    std::cerr << "    " << cloud.points[i].x << " " << cloud.points[i].y << " " << cloud.points[i].z << std::endl;

  return (0);
}

Finally you will need the following file which should be in the same folder as the sample source code.

$ vi CMakeLists.txt

cmake_minimum_required(VERSION 2.6 FATAL_ERROR)
project(pcd_write)
find_package(PCL 1.3 REQUIRED COMPONENTS common io)
include_directories(${PCL_INCLUDE_DIRS})
link_directories(${PCL_LIBRARY_DIRS})
add_definitions(${PCL_DEFINITIONS})
add_executable(pcd_write pcd_write.cpp)
target_link_libraries(pcd_write ${PCL_COMMON_LIBRARIES} ${PCL_IO_LIBRARIES})

When you have created both these files then you can finally find out if it works!

$ mkdir build
$ cd build
$ cmake ..

  :
  :
  :
  
$ make
Scanning dependencies of target pcd_write
[100%] Building CXX object CMakeFiles/pcd_write.dir/pcd_write.cpp.o
Linking CXX executable pcd_write
[100%] Built target pcd_write
$


References

  • Point Cloud Library
  • OpenNI
  • PCL/OpenNITutorial coverng the compilation, installation and testing
  • Sensor Kinect

  • This entry was posted in Hardware, Linux, Ubuntu and tagged , . Bookmark the permalink.

    Leave a Reply

    Please log in using one of these methods to post your comment:

    WordPress.com Logo

    You are commenting using your WordPress.com account. Log Out / Change )

    Twitter picture

    You are commenting using your Twitter account. Log Out / Change )

    Facebook photo

    You are commenting using your Facebook account. Log Out / Change )

    Google+ photo

    You are commenting using your Google+ account. Log Out / Change )

    Connecting to %s