Content uploaded by Nathir A Rawashdeh
Author content
All content in this area was uploaded by Nathir A Rawashdeh on May 30, 2023
Content may be subject to copyright.
A MATLAB Script for Visualizing and Processing
BOREAS Dataset Road Driving Sensor Data:
LiDAR, Radar, and Camera.
1st Dunsten M. X. Dsouza
Dept. of Applied Computing
Michigan Technological University
Houghton, MI, USA
ORCID 0009-0009-7934-7165
2nd Nader J. Abu-Alrub
Dept. of Applied Computing
Michigan Technological University
Houghton, MI, USA
ORCID 0000-0002-3930-762X
3rd Nathir A. Rawashdeh
Dept. of Applied Computing
Michigan Technological University
Houghton, MI, USA
ORCID 0000-0002-9118-9317
Abstract—This study presents an alternative approach to
accessing the BOREAS dataset by utilizing MATLAB’s built-
in functions. Recently, one of the most intriguing research areas
has been vehicle perception and simultaneous localization and
mapping (SLAM) for different weather conditions. To simplify
the process of opening the dataset in a toolbox, we have developed
this project to assist researchers. Now, instead of concentrating
on how to open the dataset in MATLAB, researchers can
concentrate on their main project work. MATLAB has impressive
functionalities for manipulating data in an offline setting. In this
work, we have used inbuilt commands to open the BOREAS
Dataset in MATLAB and inspect the various LIDAR, RADAR,
and Camera datasets for future applications. The data can be
used in real and practical experiments in the field of autonomous
driving. This thesis is divided into two parts: (i) to open the
LIDAR, RADAR, and Camera datasets with individual plots,
and (ii) to open the LIDAR, RADAR, and Camera datasets with
concurrent plots. These datasets were obtained from the Boreas:
A Multi-Season Autonomous Driving Dataset
I. INTRODUCTION
The Advanced Driver Assistance Systems (ADAS) is a
technology that can aid drivers in driving by providing ad-
ditional information, warning them of potential hazards, and
taking corrective measures in specific situations. This system
can enhance driver safety and comfort by offering additional
assistance to the driver in crucial moments. Autonomous cars
depend on various sensors to collect data about their environ-
ment and decide on how to operate the vehicle. These sensors
include cameras, lidar (light detection and ranging), radar,
and ultrasonic sensors. Cameras capture visual information
about the road, such as lane markings, traffic signals, and
other cars. Lidar and radar sensors measure the distance and
speed of the objects in the vehicle’s surroundings, including
pedestrians, other vehicles, and obstacles. Ultrasonic sensors
are used for close-range sensing, such as when parking or
manoeuvring in tight spaces. The sensors provide information
to the car’s ADAS system, which uses algorithms and artificial
intelligence to interpret the data and decide how to operate
the vehicle. For example, the ADAS system may use sensor
data to detect the road without lane markers, or prevent an
imminent collision and take evasive action as braking or
Fig. 1. Concurrent visualization of camera, liDAR, and radar
steering to avoid the collision. ADAS can also assist with
lane keeping, adaptive cruise control, and parking assistance.
Autonomous car sensors and ADAS technology provide a
safer and more efficient driving experience. ADAS technology
has the potential to reduce accidents and make driving more
accessible and enjoyable for everyone by gathering data about
the environment and assisting drivers with critical decisions.
Influence of Weather conditions on driving: This can signifi-
cantly affect driving, and it is crucial to modify your driving
behavior accordingly to ensure safety on the road. Rain, ice,
and snow can make roads slippery, reducing traction and
increasing the likelihood of accidents. Fog can also impair
visibility, challenging spotting other vehicles, pedestrians, or
road signs. It is also important to slow down and increase the
distance between vehicles, as sudden stops may be difficult to
see in low visibility conditions. In conclusion, adapting your
driving behavior according to weather conditions is necessary
to ensure safety on the road. By reducing speed, increasing the
distance between vehicles, and taking other safety measures,
you can minimize the risk of accidents and keep yourself and
others safe. Boreas has done an amazing job capturing all these
datasets.
The BOREAS dataset: This dataset [1] has been compiled by
driving along the same route repeatedly for an entire year.
It exhibits significant fluctuations due to various seasonal
and weather changes, such as precipitation and snowfall. The
dataset comprises more than 350 kilometers of driving data,
and precise ground truth postures have been post-processed
with centimetric accuracy. It is accompanied by a 128-channel
Velodyne Alpha Prime lidar, a 360-degree Navtech CIR304-H
scanning radar, and a 5-megapixel FLIR Blackfly S camera.
The dataset also provides up-to-date leaderboards for odome-
try, metric localization, and 3D object detection.
TABLE I
SEN SOR S US ED TO C OLL EC T THE BOREAS DATASE T
Sensors Used for the Boreas Data Acquisition
Sensor Name Specification
Applanix POSLV
220 Navtech
CIR304-H Radar
200Hz,0.0438 m range so-
lution,0.9° horizontal res-
olution,250m range,4 Hz
FLIR Blackfly
S,2448 ×
2048(5MP)
Camera(BFS-U3-
51S5C)
2448 × 2048 (5 MP) 81°
HFOV × 71° VFOV, 10
Hz
Velodyne Alpha-
Prim Lidar
128beams,0.1°vertical
resolu-
tion(variable),360°HFOV
×40°VFOV,300m
range(10 % reflectivity),
2.2M points/s, 10 Hz
Boreas code on Github: According to their GitHub direc-
tions, the Boreas dataset can currently be viewed only through
some Python scripts. Several scripts interface with each other
to view the LIDAR, RADAR and CAMERA Data. We have
replicated the same Python scripts using MATLABs inbuilt
functions for a better pre and post-processing of the data.
MATLAB is a well-known tool for all the functions it has, and
the AI processing for offline Data-Processing. MATLAB code
to visualize three sensor modalities: When you run the script,
the code iterates through a FOR loop for LIDAR RADAR
and CAMERA. Inside the FOR loop, some commands point
MATLAB to a directory which contains the BOREAS data. It
then iterates through these files in a synchronous order. Now
the naming convention followed for the file names is a UNIX
timestamp, which is in an ascending order. We have made use
of the time gap method to show the matching frames. When
its time to show an image, MATLAB goes through each file,
opens it, does the processing, shows the figure to the user
and then swaps the old image with a new one. This creates
a continuous visualization for each time stamp as MATLAB
iterates over these files from start to end.
Fig. 2. MATLAB visualization script pseudo code
II. MATLAB: PSEU DU CODE FOR VISUALIZATION
The above process can be enhanced by utilizing MATLAB
functions and AI technologies for offline data processing;
The RADAR, LIDAR, and VISION data on MATLAB are
imported from the BOREAS directory. The RADAR and
VISION data are images in the PNG format. But for LiDAR,
we have to process and visualize the data using the Point
Cloud Processing Toolbox. We have done the signal process-
ing, through MATLAB’s plotting tools, such as plot() and
imshow(). MATLAB is helpful for academics and engineers
working with autonomous cars or similar subjects since it
offers many tools and functions for processing and visualizing
RADAR, LIDAR, and VISION data.
III. MATLAB CODE AVAILABILITY
The MATLAB script ’Show BOREAS.m’ and some of the
sample BOREAS data files can be downloaded at Nathir
Rawashdeh’s Digital Commons page at Michigan Tech at
https://digitalcommons.mtu.edu. Extract the ZIP file, open
MATLAB, and run the script.
REFERENCES
[1] Keenan Burnett, David J Yoon, Yuchen Wu, Andrew Z Li, Haowei Zhang,
Shichen Lu, Jingxing Qian, Wei-Kang Tseng, Andrew Lambert, Keith YK
Leung, et al. Boreas: A multi-season autonomous driving dataset. The
International Journal of Robotics Research, page 02783649231160195,
2023.