readme.md

PCL Samples for Intel® RealSense™ cameras

Examples in this folder are designed to complement existing SDK examples and demonstrate how Intel RealSense cameras can be used together with PCL (Point-Cloud Library).

List of Samples:

  1. PCL - Minimal Point-cloud viewer that includes PCL processing
  2. PCL-COLOR - Point-cloud viewer that includes RGB PCL processing

Getting Started:

This page is certainly not a comprehensive guide to getting started with PCL, but it can help get on the right track.

Windows

  1. Download and install CMake from cmake.org/download
  2. Donwload prebuilt binaries for PCL (official link). For Visual Studio 2015 / 2017, use the following link.
  3. Follow the installation procedure:

  1. Open the Command Prompt and execute:
set PCL_ROOT="C:\Program Files (x86)\PCL 1.8.1"
set PATH=%PATH%;%PCL_ROOT%\bin;%OPENNI2_REDIST%

Or if you have install the 64-bit version:

set PCL_ROOT="C:\Program Files\PCL 1.8.1"
set PATH=%PATH%;%PCL_ROOT%\bin;%OPENNI2_REDIST64%
  1. If you had Visual Studio openned, please restart it to make sure new environment variables are in effect.

  2. Clone or download librealsense sources from github.com/IntelRealSense/librealsense into a local directory (C:/git/librealsense)

  3. Run cmake-gui and fill source code and binaries locations and press Configure

  4. Make sure ``BUILD_SHARED_LIBS` is checked (linking to librealsense staticly will cause problems with PCL linkage)

  5. Check the BUILD_PCL_EXAMPLES flag and click Configure again:

  1. Click Generate and Open Project
  2. Locate PCL solution-folder under Examples

  1. Right-click on one of the examples to Set as StartUp Project
  2. Press F5 to compile and run the example

Linux

  1. Install PCL from official PPA:
sudo add-apt-repository ppa:v-launchpad-jochen-sprickerhof-de/pcl
sudo apt-get update
sudo apt-get install libpcl-all
  1. Follow the instructions to build librealsense from source
  2. Add -DBUILD_PCL_EXAMPLES=true to your cmake command
  3. Run ./wrappers/pcl/pcl/rs-pcl or just rs-pcl if you have previously executed sudo make install

The demo seem to Seg-Fault in Debug configuration. This is very troubling, but it seem to be somewhat of a known issue with PCL. As a work-around add -DCMAKE_BUILD_TYPE=RelWithDebInfo to your cmake command