![Fig. 2. An example of the PCL implementation pipeline for Fast Point Feature Histogram (FPFH) [4] estimation.](/figures/fig-2-an-example-of-the-pcl-implementation-pipeline-for-fast-1osu9ug1.webp)
![Fig. 10. Experiments with PCL in grasping applications [11], from left to right: a visualization of the collision environment including points associated with unrecognized objects (blue), and obstacles with semantic information (green); detail showing 3D point cloud data (grey) with 3D meshes superimposed for recognized objects; successful grasping showing the bounding box associated with an unrecognized object (brown) attached to the gripper.](/figures/fig-10-experiments-with-pcl-in-grasping-applications-11-from-18uqa3f5.webp)
![Fig. 9. Left: example of door and handle identification [8] during a navigation and mapping experiment [9] with the PR2 robot. Right: object recognition experiments (chair, person sitting down, cart) using Normal Aligned Radial Features (NARF) [10] with the PR2 robot.](/figures/fig-9-left-example-of-door-and-handle-identification-8-k8kklo7r.webp)
3D is here: Point Cloud Library (PCL)
Radu Bogdan Rusu and Steve Cousins
Willow Garage
68 Willow Rd., Menlo Park, CA 94025, USA
{rusu,cousins}@willowgarage.com
Abstract— With the advent of new, low-cost 3D sensing
hardware such as the Kinect, and continued efforts in advanced
point cloud processing, 3D perception gains more and more
importance in robotics, as well as other fields.
In this paper we present one of our most recent initiatives in
the areas of point cloud perception: PCL (Point Cloud Library
– http://pointclouds.org). PCL presents an advanced
and extensive approach to the subject of 3D perception, and
it’s meant to provide support for all the common 3D building
blocks that applications need. The library contains state-of-
the art algorithms for: filtering, feature estimation, surface
reconstruction, registration, model fitting and segmentation.
PCL is supported by an international community of robotics
and perception researchers. We provide a brief walkthrough of
PCL including its algorithmic capabilities and implementation
strategies.
I. INTRODUCTION
For robots to work in unstructured environments, they need
to be able to perceive the world. Over the past 20 years,
we’ve come a long way, from simple range sensors based
on sonar or IR providing a few bytes of information about
the world, to ubiquitous cameras to laser scanners. In the
past few years, sensors like the Velodyne spinning LIDAR
used in the DARPA Urban Challenge and the tilting laser
scanner used on the PR2 have given us high-quality 3D
representations of the world - point clouds. Unfortunately,
these systems are expensive, costing thousands or tens of
thousands of dollars, and therefore out of the reach of many
robotics projects.
Very recently, however, 3D sensors have become available
that change the game. For example, the Kinect sensor for
the Microsoft XBox 360 game system, based on underlying
technology from PrimeSense, can be purchased for under
$150, and provides real time point clouds as well as 2D
images. As a result, we can expect that most robots in the
future will be able to ”see” the world in 3D. All that’s
needed is a mechanism for handling point clouds efficiently,
and that’s where the open source Point Cloud Library, PCL,
comes in. Figure 1 presents the logo of the project.
PCL is a comprehensive free, BSD licensed, library for
n-D Point Clouds and 3D geometry processing. PCL is
fully integrated with ROS, the Robot Operating System (see
http://ros.org), and has been already used in a variety
of projects in the robotics community.
II. ARCHITECTURE AND IMPLEMENTATION
PCL is a fully templated, modern C++ library for 3D
point cloud processing. Written with efficiency and per-
Fig. 1. The Point Cloud Library logo.
formance in mind on modern CPUs, the underlying data
structures in PCL make use of SSE optimizations heavily.
Most mathematical operations are implemented with and
based on Eigen, an open-source template library for linear
algebra [1]. In addition, PCL provides support for OpenMP
(see http://openmp.org) and Intel Threading Building
Blocks (TBB) library [2] for multi-core parallelization. The
backbone for fast k-nearest neighbor search operations is
provided by FLANN (Fast Library for Approximate Nearest
Neighbors) [3]. All the modules and algorithms in PCL pass
data around using Boost shared pointers (see Figure 2), thus
avoiding the need to re-copy data that is already present
in the system. As of version 0.6, PCL has been ported to
Windows, MacOS, and Linux, and Android ports are in the
works.
From an algorithmic perspective, PCL is meant to incor-
porate a multitude of 3D processing algorithms that operate
on point cloud data, including: filtering, feature estimation,
surface reconstruction, model fitting, segmentation, registra-
tion, etc. Each set of algorithms is defined via base classes
that attempt to integrate all the common functionality used
throughout the entire pipeline, thus keeping the implementa-
tions of the actual algorithms compact and clean. The basic
interface for such a processing pipeline in PCL is:
• create the processing object (e.g., filter, feature estima-
tor, segmentation);
• use setInputCloud to pass the input point cloud dataset
to the processing module;
• set some parameters;
• call compute (or filter, segment, etc) to get the output.
The sequence of pseudo-code presented in Figure 2 shows
a standard feature estimation process in two steps, where a
NormalEstimation object is first created and passed an input
dataset, and the results together with the original input are
then passed together to an FPFH [4] estimation object.
To further simplify development, PCL is split into a series
of smaller code libraries, that can be compiled separately:
• libpcl filters: implements data filters such as downsam-
NormalEstimation
PointCloudConstSharedPtr &cloud
[PointCloud &normals]
PointCloudConstSharedPtr &normals
FPFHEstimation
[PointCloud &fpfh]
Fig. 2. An example of the PCL implementation pipeline for Fast Point
Feature Histogram (FPFH) [4] estimation.
pling, outlier removal, indices extraction, projections,
etc;
• libpcl features: implements many 3D features such as
surface normals and curvatures, boundary point estima-
tion, moment invariants, principal curvatures, PFH and
FPFH descriptors, spin images, integral images, NARF
descriptors, RIFT, RSD, VFH, SIFT on intensity data,
etc;
• libpcl io: implements I/O operations such as writing
to/reading from PCD (Point Cloud Data) files;
• libpcl segmentation: implements cluster extraction,
model fitting via sample consensus methods for a va-
riety of parametric models (planes, cylinders, spheres,
lines, etc), polygonal prism extraction, etc’
• libpcl surface: implements surface reconstruction tech-
niques, meshing, convex hulls, Moving Least Squares,
etc;
• libpcl registration: implements point cloud registration
methods such as ICP, etc;
• libpcl keypoints: implements different keypoint extrac-
tion methods, that can be used as a preprocessing step
to decide where to extract feature descriptors;
• libpcl range image: implements support for range im-
ages created from point cloud datasets.
To ensure the correctness of operations in PCL, the
methods and classes in each of the above mentioned libraries
contain unit and regression tests. The suite of unit tests is
compiled on demand and verified frequently by a dedicated
build farm, and the respective authors of a specific compo-
nent are being informed immediately when that component
fails to test. This ensures that any changes in the code are
tested throughly and any new functionality or modification
will not break already existing code that depends on PCL.
In addition, a large number of examples and tutorials
are available either as C++ source files, or as step-by-step
instructions on the PCL wiki web pages.
III. PCL AND ROS
One of the corner stones in the PCL design philosophy
is represented by Perception Processing Graphs (PPG). The
rationality behind PPGs are that most applications that deal
with point cloud processing can be formulated as a concrete
set of building blocks that are parameterized to achieve dif-
ferent results. For example, there is no algorithmic difference
between a wall detection algorithm, or a door detection, or a
table detection – all of them share the same building block,
which is in this case, a constrained planar segmentation
algorithm. What changes in the above mentioned cases is
a subset of the parameters used to run the algorithm.
With this in mind, and based on the previous experience of
designing other 3D processing libraries, and most recently,
ROS, we decided to make each algorithm from PCL available
as a standalone building block, that can be easily connected
with other blocks, thus creating processing graphs, in the
same way that nodes connect together in a ROS ecosystem.
Furthermore, because point clouds are extremely large in
nature, we wanted to guarantee that there would be no
unnecessary data copying or serialization/deserialization for
critical applications that can afford to run in the same
process. For this we created nodelets, which are dynamically
loadable plugins that look and operate like ROS nodes, but
in a single process (as single or multiple threads).
A concrete nodelet PPG example for the problem of
identifying a set of point clusters supported by horizontal
planar areas is shown in Figure 3.
Fig. 3. A ROS nodelet graph for the problem of object clustering on planar
surfaces.
IV. VISUALIZATION
PCL comes with its own visualization library, based on
VTK [5]. VTK offers great multi-platform support for ren-
dering 3D point cloud and surface data, including visualiza-
tion support for tensors, texturing, and volumetric methods.
The PCL Visualization library is meant to integrate PCL
with VTK, by providing a comprehensive visualization layer
for n-D point cloud structures. Its purpose is to be able
to quickly prototype and visualize the results of algorithms
operating on such hyper-dimensional data. As of version 0.2,
the visualization library offers:
• methods for rendering and setting visual properties
(colors, point sizes, opacity, etc) for any n-D point cloud
dataset;
• methods for drawing basic 3D shapes on screen (e.g.,
cylinders, spheres, lines, polygons, etc) either from sets
of points or from parametric equations;
• a histogram visualization module (PCLHistogramVisu-
alizer) for 2D plots;
• a multitude of geometry and color handlers. Here, the
user can specify what dimensions are to be used for the
point positions in a 3D Cartesian space (see Figure 4),
or what colors should be used to render the points (see
Figure 5);
• RangeImage visualization modules (see Figure 6).
The handler interactors are modules that describe how
colors and the 3D geometry at each point in space are
computed, displayed on screen, and how the user interacts
with the data. They are designed with simplicity in mind,
and are easily extendable. A code snippet that produces
results similar to the ones shown in Figure 4 is presented
in Algorithm 1.
Algorithm 1 Code example for the results shown in Figure 4.
using namespace pcl visualization;
PCLVisualizer p (“Test”);
PointCloudColorHandlerRandom handler (cloud);
p.addPointCloud (cloud, handler, ”cloud random”);
p.spin ();
p.removePointCloud (”cloud random”);
PointCloudGeometryHandlerSurfaceNormal handler2 (cloud);
p.addPointCloud (cloud, handler2, ”cloud random”);
p.spin ();
The library also offers a few general purpose tools for
visualizing PCD files, as well as for visualizing streams of
data from a sensor in real-time in ROS.
Fig. 4. An example of two different geometry handers applied to the same
dataset. Left: the 3D Cartesian space represents XYZ data, with the arrows
representing surface normals estimated at each point in the cloud, right: the
Cartesian space represents the 3 dimensions of the normal vector at each
point for the same dataset.
Fig. 5. An example of two different color handers applied to the
same dataset. Left: the colors represent the distance from the acquisition
viewpoint, right: the color represent the RGB texture acquired at each point.
Fig. 6. An example of a RangeImage display using PCL Visualization
(bottom) for a given 3D point cloud dataset (top).
V. USAGE EXAMPLES
In this section we present two code snippets that exhibit
the flexibility and simplicity of using PCL for filtering
and segmentation operations, followed by three application
examples that make use of PCL for solving the perception
problem: i) navigation and mapping, ii) object recognition,
and iii) manipulation and grasping.
Filtering constitutes one of the most important operations
that any raw point cloud dataset usually goes through, before
any higher level operations are applied to it. Algorithm 2 and
Figure 7 present a code snippet and the results obtained after
running it on the point cloud dataset from the left part of the
figure. The filter is based on estimating a set of statistics for
the points in a given neighborhood (k = 50 here), and using
them to select all points within 1.0· σ distance from the mean
distance µ, as inliers (see [6] for more information).
Algorithm 2 Code example for the results shown in Figure 7.
pcl::StatisticalOutlierRemoval<pcl::PointXYZ> f;
f.setInputCloud (input cloud);
f.setMeanK (50);
f.setStddevMulThresh (1.0);
f.filter (output cloud);
Fig. 7. Left: a raw point cloud acquired using a tilting laser scanner,
middle: the resultant filtered point cloud (i.e., inliers) after a StatisticalOut-
lierRemoval operator was applied, right: the rejected points (i.e., outliers).
The second example constitutes a segmentation operation
for planar surfaces, using a RANSAC [7] model, as shown
in Algorithm 3. The input and output results are shown in
Figure 8. In this example, we are using a robust RANSAC
estimator to randomly select 3 non-collinear points and
calculate the best possible model in terms of the overall
number of inliers. The inlier thresholding criterion is set to a
maximum distance of 1cm of each point to the plane model.
Algorithm 3 Code example for the results shown in Figure 8.
pcl::SACSegmentation<pcl::PointXYZ> s;
f.setInputCloud (input cloud);
f.setModelType (pcl::SACMODEL PLANE);
f.setMethodType (pcl::SAC RANSAC);
f.setDistanceThreshold (0.01);
f.segment (output cloud);
Fig. 8. Left: the input point cloud, right: the segmented plane represented
by the inliers of the model marked with purple color.
An example of a more complex navigation and mapping
application is shown in the left part of Figure 9, where the
PR2 robot had to autonomously identify doors and their
handles [8], in order to explore rooms and find power
sockets [9]. Here, the modules used included constrained
planar segmentation, region growing methods, convex hull
estimation, and polygonal prism extractions. The results of
these methods were then used to extract certain statistics
about the shape and size of the door and the handle, in order
to uniquely identify them and to reject false positives.
The right part of Figure 9 shows an experiment with
real-time object identification from complex 3D scenes [10].
Here, a set of complex 3D keypoints and feature descriptors
are used in a segmentation and registration framework, that
aims to identify previously seen objects in the world.
Figure 10 presents a grasping and manipulation applica-
tion [11], where objects are first segmented from horizontal
planar tables, clustered into individual units, and a registra-
tion operation is applied that attaches semantic information
to each cluster found.
VI. COMMUNITY AND FUTURE PLANS
PCL is a large collaborative effort, and it would not
exist without the contributions of several people. Though
the community is larger, and we accept patches and im-
provements from many users, we would like to acknowledge
the following institutions for their core contributions to the
development of the library: AIST, UC Berkeley, University of
Handle
Door
Fig. 9. Left: example of door and handle identification [8] during a
navigation and mapping experiment [9] with the PR2 robot. Right: object
recognition experiments (chair, person sitting down, cart) using Normal
Aligned Radial Features (NARF) [10] with the PR2 robot.
Fig. 10. Experiments with PCL in grasping applications [11], from
left to right: a visualization of the collision environment including points
associated with unrecognized objects (blue), and obstacles with semantic
information (green); detail showing 3D point cloud data (grey) with 3D
meshes superimposed for recognized objects; successful grasping showing
the bounding box associated with an unrecognized object (brown) attached
to the gripper.
Bonn, University of British Columbia, ETH Zurich, Univer-
sity of Freiburg, Intel Reseach Seattle, LAAS/CNRS, MIT,
University of Osnabr
¨
uck, Stanford University, University of
Tokyo, TUM, Vienna University of Technolog, and Wash-
ington University in St. Louis.
Our current plan for PCL is to improve the documentation,
unit tests, and tutorials and release a 1.0 version. We will
continue to add functionality and make the system available
on other platforms such as Android, and we plan to add
support for GPUs using CUDA and OpenCL.
We welcome any new contributors to the project, and
we hope to emphasize the importance of code sharing for
3D processing, which is becoming crucial for advancing the
robotics field.
REFERENCES
[1] G. Guennebaud, B. Jacob, et al., “Eigen v3,” http://eigen.tuxfamily.org,
2010.
[2] J. Reinders, Intel threading building blocks : outfitting C++ for multi-
core processor parallelism. O’Reilly, 2007.
[3] M. Muja and D. G. Lowe, “Fast approximate nearest neighbors with
automatic algorithm configuration,” in International Conference on
Computer Vision Theory and Application VISSAPP’09). INSTICC
Press, 2009, pp. 331–340.
[4] R. B. Rusu, N. Blodow, and M. Beetz, “Fast Point Feature Histograms
(FPFH) for 3D Registration,” in Proceedings of the IEEE International
Conference on Robotics and Automation (ICRA), Kobe, Japan, May
12-17 2009.
[5] W. Schroeder, K. Martin, and B. Lorensen, Visualization Toolkit: An
Object-Oriented Approach to 3D Graphics, 4th Edition. Kitware,
December 2006.
[6] R. B. Rusu, Z. C. Marton, N. Blodow, M. Dolha, and M. Beetz,
“Towards 3D Point Cloud Based Object Maps for Household Envi-
ronments,” Robotics and Autonomous Systems Journal (Special Issue
on Semantic Knowledge), 2008.
[7] A. M. Fischler and C. R. Bolles, “Random sample consensus: a
paradigm for model fitting with applications to image analysis and
automated cartography,” Communications of the ACM, vol. 24, no. 6,
pp. 381–395, June 1981.
[8] R. B. Rusu, W. Meeussen, S. Chitta, and M. Beetz, “Laser-based
Perception for Door and Handle Identification,” in International Con-
ference on Advanced Robotics (ICAR), June 22-26 2009.
[9] W. Meeussen, M. Wise, S. Glaser, S. Chitta, C. McGann, P. Mihelich,
E. Marder-Eppstein, M. Muja, V. Eruhimov, T. Foote, J. Hsu, R. Rusu,
B. Marthi, G. Bradski, K. Konolige, B. Gerkey, and E. Berger,
“Autonomous Door Opening and Plugging In with a Personal Robot,”
in Proceedings of the IEEE International Conference on Robotics and
Automation (ICRA), Anchorage, Alaska, May 3-8 2010.
[10] B. Steder, R. B. Rusu, K. Konolige, and W. Burgard, “Point Feature
Extraction on 3D Range Scans Taking into Account Object Bound-
aries,” in Submitted to the IEEE International Conference on Robotics
and Automation (ICRA), Shanghai, China, May 9-13 2010.
[11] M. Ciocarlie, K. Hsiao, E. G. Jones, S. Chitta, R. B. Rusu, and
I. A. Sucan, “Towards reliable grasping and manipulation in household
environments,” New Delhi, India, 12/2010 2010.