Direct Georeferencing 3D Points Cloud Map Based

on SLAM and Robot Operating System

*

Fraj Hariz Haifa Souifi 3rd Ryan LeBlanc

Electrical Engineering Department Electrical Engineering Department Electrical Engineering Department
Université de Moncton Université de Moncton Université de Moncton
Moncton, Canada Moncton, Canada Moncton, Canada
efh5474@umoncton.ca ehs4973@umoncton.ca ryan.leblanc@umoncton.ca

Yassine Bouslimani Mohsen Ghribi Eric Langin

Electrical Engineering Department Electrical Engineering Department Eagle GIS
Université de Moncton Université de Moncton Fredericton, Canada
Moncton, Canada Moncton, Canada eric.langin@eagle-gis.com
yassine.bouslimani@umoncton.ca mohsen.ghribi@umoncton.ca

Dan Mccarthy
Eagle Telecom
Fredericton, Canada
dan.mccarthy@eagletelecom.ca

Abstract—Nowadays, the 3D data has known an explosive

demand by many applications such as: creation of 3D maps,
road condition monitoring and autonomous navigation. The
concept of collecting geospatial data from the road environment
has recognized by the term of «Mobile Mapping» which
represents a cost efficient and quick method to acquire data
from dynamic fields such as highways, underground streets,
and urban zones, in where the process of data collection has
known a difficulty of access due to the road traffic or to bad
weather conditions. Traditional mobile mapping systems can
face several problems regard possibly the reliability issues and
poor accuracy of produced data, particularly in the GNSS
Fig. 1. Components of Mobile Mapping System
outage case. For this reason, the integration of Simultaneous
Localization and Mapping (SLAM) method in Mobile Mapping
System (MMS) represents a robust mapping solution in com-
plex environments to optimize the accuracy of the generated to collect geographic data from the road environment by
data even in the case of GNSS signal loss. The present paper combining several types of data such as: GNSS global
proposes an efficient approach for collecting, modeling, and navigation data, Inertial Navigation (IMU) data, 3D
generating large-scale 3D point cloud maps in real-world coordi- LiDAR data, and panoramic images.
nates based on SLAM algorithm and Robot Operating System
(ROS). The proposed approach can effectively generate a high-
definition georeferenced point cloud maps with an accuracy up
Although the existing MMS on the market have many
to ± 5 cm. advantages in terms of speed and ease of use, their high
Keywords—SLAM, ROS, Direct Georeferencing, LiDAR 3D, cost, roughly up to 2 million $ USD, still represents a
GIS, MMS, GNSS/INS big challenge for developing new mapping applications
[2]. Furthermore, their main weakness appears essentially
I. Introduction during the geospatial data collection, in where each 3D
Typically, a Mobile Mapping System (MMS) is com- point cloud heavily depends on the previous point cloud
posed of four main elements which are: (1) Positioning and the current GPS position as well [3]. Therefore, these
sensors (GNSS, IMU), (2) Measurement sensors, (3) On- systems can face several problems regard possibly the
board computer with large data storage space, and (4) reliability issues and poor accuracy of produced data,
Mobile vehicle (Fig. 1) [1]. Its fundamental function is particularly in the GNSS outage case.

Over the last two decades, many lidar-based methods

have become increasingly attractive for almost mapping To this end, this study aims to develop a low-cost mobile
applications owing to the reliability of 3D LiDAR data mapping system that produces georeferenced 3D point
compared to vision-based on photogrammetry techniques. cloud maps using SLAM method and ROS. The main
As stated in the literature [4, 5], a wide range of applica- contributions of this investigation can be summarized as
tions require more and more denser and more accurate 3D follows:
maps. This kind of map was widely known as the High- 1) Design the mobile mapping system that integrates
Definition Maps (HD Maps). In fact, providing HD map a 3D LiDAR sensor, an inertial measurement unit
would be helpful for the smoothest development of inno- (IMU), and GNSS receiver based on high-accuracy
vative applications including the autonomous driving, the technology.
construction of smart city, detection and 3D modelization 2) Integration and synchronization of different hard-
of objects [4][5]. ware technologies.
3) Developing the required software to collect geospa-
In this context, tremendous efforts have been recently tial data from complex environment using ROS.
put on proposing techniques to combine several sources of 4) (4) Proposing an efficient approach based on SLAM
the collected data (data fusion) in order to construct 3D that produces georeferenced 3D point cloud maps in
map for an unknown environment [6]. This process was format of most popular GIS files such as *.las, *.pcd,
commonly called SLAM (Simultaneous Localization and and *.shp files.
Mapping [6].
The remainder of this paper is presented in 3 sections.
The most common SLAM method is the Cartographer Section II presents the proposed mobile mapping system.
ROS developed by Google, which is an open-source soft- In Section III, a new approach for generating georeferenced
ware that provides real-time simultaneous localization and point cloud map was explored using a schematic diagram.
mapping in 2D or 3D modes [7]. So as to optimize the In Section IV, all experiments and results are reported.
accuracy of the typical SLAM methods, many SLAM Finally, Section V addresses the major conclusions of the
algorithms are born and have evolved continuously. present paper.

In [8], the authors developed a new strategy of SLAM by II. SYSTEM DESIGN AND SETUP
using a LiDAR sensor and an inertial measurement unit A. Selecting the Appropriate Hardware
(IMU). The proposed strategy was to divide the simulta- Selecting an appropriate hardware that can be used in
neous localization and mapping process by two algorithms. a new mobile mapping system was the first challenge. It
The first algorithm is dedicated to determining the pose is undeniable that the positioning system is one of the
of the robot by providing the odometry data. Another cornerstones of the new design.
algorithm is used for matching and registration point cloud
based on Iterative Closest Point (ICP algorithm). For this reason, the intended study focuses on acquiring
an existing positioning system that can combine inertial
Despite the high performance of the LOAM algorithm,
data (IMU) and GNSS receiver in order to meet the level
its inability to integrate GPS data for pose correction as
of accuracy needed by almost GIS applications and to
well as its instability during implementation in outdoor
adapt with the vehicle speeds as well. In this respect, the
mapping applications still represent its big issue.
PwrPak7D-E2 enclosure was considered as the core of the
To overcome these limitations, Shan et al. [9] shed light proposed design.
on an alternative approach named (Tightly coupled Lidar
Inertial Odometry via Smoothing and Mapping) which Owing to the SPAN technology, this positioning system
requires an IMU with 9-axis (3 accelerometers, 3 gyro- includes an Epson G370 MEMS Inertial Measurement
scopes, and 3 magnetometers) to provide the odometry Unit (IMU) to deliver a geographic position each 5 ms
data. LIO-SAM achieves highly accurate, real-time mobile (Data Rate up to 200 Hz). PwrPak7D-E2 with dual GNSS
robot trajectory estimation and map-building thanks to antenna allows reaching the best accuracy provided by
the integration of multiple sources of data like GPS Novatel Correction Services. The main performances and
measurements, IMU data, and lidar frames and to the limitations of the proposed positioning system for each
construction of a global map via multiple parts. service were presented in Table 1.
The IEEEtran class file is used to format your paper
SLAM algorithms produce a local map, in which the and style the text. All margins, column widths, line spaces,
reference coordinate system is the robot frame and the and text fonts are prescribed; please do not alter them.
map’s origin (0,0,0) represents the initial position of robot You may note peculiarities. For example, the head margin
before executing SLAM process. As a matter of fact, measures proportionately more than is customary. This
SLAM only methods cannot deal with the real demand of measurement and others are deliberate, using specifica-
almost GIS application which require georeferenced data. tions that anticipate your paper as one part of the entire
 TABLE I symbols. Use a long dash rather than a hyphen for a
TABLE I. NOVATEL CORRECTION SERVICES minus sign. Punctuate equations with commas or periods
CHARACTERISTICS
when they are part of a sentence, as in:
Novatel Accuracy Convergence
Weakness
Correction Service GNSS Time
GNSS
Outage a+b=γ (1)
Complex architecture composed of
base and rover stations.

RTK 0.01 m 0.02 – No initial

Expensive solution.
Difficulty to establish a
Be sure that the symbols in your equation have been
5.02 m convergence
period.
communication link between
base and rover stations for
defined before or immediately following the equation. Use
TerraStar-L 0.4 m 5 min
big distance.
Poor Accuracy
“(1)”, not “Eq. (1)” or “equation (1)”, except at the
TerraStar-C
TerraStar-C PRO
0.04 m
0.025 m
30 min
0.06 m – 5.06 m
18 min
beginning of a sentence: “Equation (1) is . . .”
Long waiting time to obtain the converged solution.
TerraStar-X 0.025 m 1 min Coverage zone is very limited by this
service.
D. LATEX-Specific Advice
Please use “soft” (e.g., \eqref{Eq}) cross references
proceedings, and not as an independent document. Please instead of “hard” references (e.g., (1)). That will make it
do not revise any of the current designations. possible to combine sections, add equations, or change the
order of figures or citations without having to go through
III. Prepare Your Paper Before Styling
the file line by line.
Before you begin to format your paper, first write and Please don’t use the {eqnarray} equation environment.
save the content as a separate text file. Complete all con- Use {align} or {IEEEeqnarray} instead. The {eqnarray}
tent and organizational editing before formatting. Please environment leaves unsightly spaces around relation sym-
note sections III-A–III-E below for more information on bols.
proofreading, spelling and grammar. Please note that the {subequations} environment in
Keep your text and graphic files separate until after the LATEX will increment the main equation counter even when
text has been formatted and styled. Do not number text there are no equation numbers displayed. If you forget
heads—LATEX will do that for you. that, you might write an article in which the equation
A. Abbreviations and Acronyms numbers skip from (17) to (20), causing the copy editors
Define abbreviations and acronyms the first time they to wonder if you’ve discovered a new method of counting.
are used in the text, even after they have been defined BibTEX does not work by magic. It doesn’t get the
in the abstract. Abbreviations such as IEEE, SI, MKS, bibliographic data from thin air but from .bib files. If
CGS, ac, dc, and rms do not have to be defined. Do not you use BibTEX to produce a bibliography you must send
use abbreviations in the title or heads unless they are the .bib files.
unavoidable. LATEX can’t read your mind. If you assign the same label
to a subsubsection and a table, you might find that Table
B. Units I has been cross referenced as Table IV-B3.
• Use either SI (MKS) or CGS as primary units. (SI LATEX does not have precognitive abilities. If you put
units are encouraged.) English units may be used as a \label command before the command that updates the
secondary units (in parentheses). An exception would counter it’s supposed to be using, the label will pick up the
be the use of English units as identifiers in trade, such last counter to be cross referenced instead. In particular,
as “3.5-inch disk drive”. a \label command should not go before the caption of a
• Avoid combining SI and CGS units, such as current figure or a table.
in amperes and magnetic field in oersteds. This often Do not use \nonumber inside the {array} environment.
leads to confusion because equations do not balance It will not stop equation numbers inside {array} (there
dimensionally. If you must use mixed units, clearly won’t be any anyway) and it might stop a wanted equation
state the units for each quantity that you use in an number in the surrounding equation.
equation.
E. Some Common Mistakes
• Do not mix complete spellings and abbreviations of
units: “Wb/m2 ” or “webers per square meter”, not • The word “data” is plural, not singular.
“webers/m2 ”. Spell out units when they appear in • The subscript for the permeability of vacuum µ0 ,
text: “. . . a few henries”, not “. . . a few H”. and other common scientific constants, is zero with
• Use a zero before decimal points: “0.25”, not “.25”. subscript formatting, not a lowercase letter “o”.
Use “cm3 ”, not “cc”.) • In American English, commas, semicolons, periods,
question and exclamation marks are located within
C. Equations quotation marks only when a complete thought or
Number equations consecutively. To make your equa- name is cited, such as a title or full quotation.
tions more compact, you may use the solidus ( / ), the When quotation marks are used, instead of a bold
exp function, or appropriate exponents. Italicize Roman or italic typeface, to highlight a word or phrase,
symbols for quantities and variables, but not Greek punctuation should appear outside of the quotation
 marks. A parenthetical phrase or statement at the should be used and, conversely, if there are not at least
end of a sentence is punctuated outside of the closing two sub-topics, then no subheads should be introduced.
parenthesis (like this). (A parenthetical sentence is
punctuated within the parentheses.) H. Figures and Tables
• A graph within a graph is an “inset”, not an “insert”. a) Positioning Figures and Tables: Place figures and
The word alternatively is preferred to the word tables at the top and bottom of columns. Avoid placing
“alternately” (unless you really mean something that them in the middle of columns. Large figures and tables
alternates). may span across both columns. Figure captions should be
• Do not use the word “essentially” to mean “approxi- below the figures; table heads should appear above the
mately” or “effectively”. tables. Insert figures and tables after they are cited in the
• In your paper title, if the words “that uses” can text. Use the abbreviation “Fig. 2”, even at the beginning
accurately replace the word “using”, capitalize the of a sentence.
“u”; if not, keep using lower-cased.
• Be aware of the different meanings of the homophones TABLE II
Table Type Styles
“affect” and “effect”, “complement” and “compli-
ment”, “discreet” and “discrete”, “principal” and
“principle”.
• Do not confuse “imply” and “infer”.
• The prefix “non” is not a word; it should be joined
to the word it modifies, usually without a hyphen.
• There is no period after the “et” in the Latin
abbreviation “et al.”.
• The abbreviation “i.e.” means “that is”, and the
abbreviation “e.g.” means “for example”.
An excellent style manual for science writers is [7].
Fig. 2. Example of a figure caption.
F. Authors and Affiliations
The class file is designed for, but not limited to, six Figure Labels: Use 8 point Times New Roman for Figure
authors. A minimum of one author is required for all labels. Use words rather than symbols or abbreviations
conference articles. Author names should be listed starting when writing Figure axis labels to avoid confusing the
from left to right and then moving down to the next reader. As an example, write the quantity “Magnetiza-
line. This is the author sequence that will be used in tion”, or “Magnetization, M”, not just “M”. If including
future citations and by indexing services. Names should units in the label, present them within parentheses. Do
not be listed in columns nor group by affiliation. Please not label axes only with units. In the example, write
keep your affiliations as succinct as possible (for example, “Magnetization (A/m)” or “Magnetization {A[m(1)]}”,
do not differentiate among departments of the same not just “A/m”. Do not label axes with a ratio of quantities
organization). and units. For example, write “Temperature (K)”, not
“Temperature/K”.
G. Identify the Headings
Headings, or heads, are organizational devices that Acknowledgment
guide the reader through your paper. There are two types: The preferred spelling of the word “acknowledgment” in
component heads and text heads. America is without an “e” after the “g”. Avoid the stilted
Component heads identify the different components of expression “one of us (R. B. G.) thanks . . .”. Instead, try
your paper and are not topically subordinate to each other. “R. B. G. thanks. . .”. Put sponsor acknowledgments in the
Examples include Acknowledgments and References and, unnumbered footnote on the first page.
for these, the correct style to use is “Heading 5”. Use
“figure caption” for your Figure captions, and “table head” References
for your table title. Run-in heads, such as “Abstract”, Please number citations consecutively within brackets
will require you to apply a style (in this case, italic) in [1]. The sentence punctuation follows the bracket [2]. Refer
addition to the style provided by the drop down menu to simply to the reference number, as in [3]—do not use “Ref.
differentiate the head from the text. [3]” or “reference [3]” except at the beginning of a sentence:
Text heads organize the topics on a relational, hierar- “Reference [3] was the first . . .”
chical basis. For example, the paper title is the primary Number footnotes separately in superscripts. Place the
text head because all subsequent material relates and actual footnote at the bottom of the column in which it
elaborates on this one topic. If there are two or more sub- was cited. Do not put footnotes in the abstract or reference
topics, the next level head (uppercase Roman numerals) list. Use letters for table footnotes.
 Unless there are six authors or more give all authors’
names; do not use “et al.”. Papers that have not been
published, even if they have been submitted for publica-
tion, should be cited as “unpublished” [4]. Papers that
have been accepted for publication should be cited as “in
press” [5]. Capitalize only the first word in a paper title,
except for proper nouns and element symbols.
For papers published in translation journals, please give
the English citation first, followed by the original foreign-
language citation [6].
References
[1] G. Eason, B. Noble, and I. N. Sneddon, “On certain integrals of
Lipschitz-Hankel type involving products of Bessel functions,”
Phil. Trans. Roy. Soc. London, vol. A247, pp. 529–551, April
1955.
[2] J. Clerk Maxwell, A Treatise on Electricity and Magnetism, 3rd
ed., vol. 2. Oxford: Clarendon, 1892, pp.68–73.
[3] I. S. Jacobs and C. P. Bean, “Fine particles, thin films and
exchange anisotropy,” in Magnetism, vol. III, G. T. Rado and
H. Suhl, Eds. New York: Academic, 1963, pp. 271–350.
[4] K. Elissa, “Title of paper if known,” unpublished.
[5] R. Nicole, “Title of paper with only first word capitalized,” J.
Name Stand. Abbrev., in press.
[6] Y. Yorozu, M. Hirano, K. Oka, and Y. Tagawa, “Electron
spectroscopy studies on magneto-optical media and plastic
substrate interface,” IEEE Transl. J. Magn. Japan, vol. 2, pp.
740–741, August 1987 [Digests 9th Annual Conf. Magnetics
Japan, p. 301, 1982].
[7] M. Young, The Technical Writer’s Handbook. Mill Valley, CA:
University Science, 1989.

IEEE conference templates contain guidance text for com-

posing and formatting conference papers. Please ensure
that all template text is removed from your conference
paper prior to submission to the conference. Failure to
remove the template text from your paper may result in
your paper not being published.