Volume 4, Issue 12, December – 2019 International Journal of Innovative Science and Research Technology

ISSN No:-2456-2165

Implementation of Potholes Detection System Using

Nanodrone and Image Processing
1
Chinmai Mysorekar, 2Lijo Jose Puthenpurayil, 3Rohan Moraes, 4Amit Thool, 5Gaurav Singh
1,2,3,4,5
Dalvik Apps, Mumbai, India

Abstract:- Aging, poor maintenance and increased in the improper maintenance of roads on regular basis, friction
number of vehicles has linearly increased the number of because of vehicles leading to cracks on roads are some of
potholes and humps on the roads. In order to ensure safe the reasons.
transportation, we have designed a prototype for
potholes and humps detection. This paper aims to Potholes, Humps, Speed-Breakers has a huge
develop a technology that detects abnormality on the economic impact and hurdles for daily commute. Few
road using Unmanned Aerial Vehicles (UAVs). This approaches have been made to these hurdles including
information will be stored and updated continuously on throw-and-go method, In-Car Artificial Intelligence (AI),
a web server using the concept of Internet of Things use of Microwaves and Magnets, etc. Now a days, studies of
(IoT). Such critical road quality information can be highly automatic detection is widely used in many regions.
made public or shared with local authorities and Identification of potholes is not enough but repairing it can
government bodies for corrective action and help the commute to be smoother. Automatic systems not
maintenance of road. only help in reducing human efforts but also helps to save
time, also, it is more efficient.[9,6]
Keywords:- Potholes, Humps, Unmanned Aerial Vehicles,
Internet Of Things, Web Servers, Roads. This paper focuses on detection and identification on
potholes, humps and speed breakers which on detection feed
I. INTRODUCTION the information to the system whose access can be given to
the road civil engineering department. Use of this
Development of the country depends on the automation system can be a boon to solve our day-to-day
population, industrialization and infrastructure. Road problems and make commute safer and smoother.
conditions are the main concern for development of the
infrastructure. Aged and poorly maintained roads degrade III. PROBLEM STATEMENT
the quality road as well as the development, which increases
the need for inspection and good maintenance of the roads. Using Black-Box Camera, a new pothole-detection
Unseen Potholes, Speed breakers, Humps are one of the system using a commercial black-box camera. The proposed
main reasons of road accidents. According to the Road system detects potholes over a wide area and at low cost.
Accident Report of India, 2017, there were 3597 deaths Where they have developed a novel pothole-detection
because of Potholes and around 3248 because of Humps on algorithm specifically designed to work with the embedded
roads.[13] computing environments of black-box cameras. A pothole-
detection algorithm is installed on an embedded board in the
To reduce the damage and loss of lives, many traffic black-box camera. This algorithm collects information
system equipment’s are invented and studied which contains regarding the size of potholes and their location, and this
laser detections, smart traffic control, RADAR, Ultrasonic information is stored in the black box and then transmitted to
Sensors and many more. Unmanned Aerial Vehicle (UAV) a pothole-management server.[2]
most likely called as Drones can be used to modernize the
Road Safety Methods and hence improve the infrastructure Another automatic pothole detection system which
of the country. uses laser technology. The proposed system captures the
geographical location coordinates of the potholes and
A model of pothole detection is constructed using the image humps. It can actively learn the knowledge about the
library, which is used in an algorithmic approach that suspension system of the host vehicle without any human
combines a road color model with simple image processing intervention vibration model to infer the presence of pothole
techniques such as a Canny filter and contour detection. while the vehicle is hitting the pothole. To ensure the use of
Using this approach, it is possible to detect potholes with a laser technology, they have used raspberry pi for software-
higher percentage of precision. based control over the system.[3]

II. PROBLEM STATEMENT Drones are used to detect potholes over the roads using
Real Time Dense Stereo Embedded System. The major
Abnormality in the roads has caused many mishaps, drawback for this proposed paper was, it was expensive,
leading to cost of lives which becomes a huddle to commute Manual visual inspection can increase human errors, time
safely on roads. The abnormality can occur because of consuming and also dangerous for humans to fly such heavy
varied different reasons. Ground water, aging of roads, drones with such high tech and heavy components.[4]

IJISRT19DEC582 www.ijisrt.com 567

Volume 4, Issue 12, December – 2019 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165

Images of the roads were acquired by Unmanned

Aerial Vehicle (UAV) and four popular supervised learning
algorithms (KNN, SVM, ANN, RF) were used to distinguish
between the normal pavement and pavement damages (i.e.
cracks and potholes). Each of learning algorithms was given
a series of different parameters, and the classification
accuracy and computational time as two assessment criteria
of the algorithm performance were calculated.[5]

IV. PROPOSED METHODOLOGY

Detecting real world images using UAV, Pluto X,

which can be used to get real time data and images from a Fig 2:- Reception Section
definite height without actually bumping through the
potholes or humps. The camera is pre mounted on the Pluto VI. COMPONENTS USED
X drone, whose output is clear and communicated through
the ESP12F Wi-Fi module integrated in the Embedded board  Hardware Components used;
which acts as a control processing unit for the UAV.
 Pluto X.
This real-world image is not only captured but is sent UAVs are commonly used where there is a risk of
to the computer using wireless communication (Wi-Fi). This sending human piloted aircrafts, or the situation makes
captured image undergoes grey scaling in image processing going of human pilot impractical. In modern times, drones
where it is scaled and externals factors are filtered out to are gaining popularity in day to day applications, which we
make it easier for the machine to detect potholes and humps. have also considered while designing this prototype.
Python can be used for scaling which later undergoes
blurring and morphology, helping to detect the edge of the Pluto X is one of the most advanced and practical
roads. drones which can be used in many applications, having a net
weight of 60 grams, makes it portable and has high
This method is used to detect potholes and humps on definition camera integrated. This is the only stand-alone
the road. This data is then saved in server, to help drivers hardware used because it is smart and designed to perform
access and get information about the road conditions for the multi-tasks. It also has 4 inbuild sensors, namely
way they have selected. This data can also be sent to the Accelerometer, Barometer, Gyro sensor and Magnetometer,
civil department to notify them and help them to understand which helps to drone to maintain its stability and gain
the exact situations of the roads and repair it on time-to-time information faster and smoother.
basis.
Inbuild Wi-Fi module, ESP12F, makes it easier to send
V. NETWORK ARCHITECTURE data to the computer wirelessly. This transfer is fast and
designed such that it is pre-processed to remove the air
 Transmission Section filters and get a clear image.
Transmission section is nothing but portable device as
UAV with camera and Wi-Fi module for transmitting data.  Technical specifications:

Processor: STM32F303: 72Mhz

Sensor suite: 10 DOF
Weight with battery: 60 g
Max Range: 60 m
Propulsion: Brushed
Photo Format: JPEG
Photo Resolution: 72ppi
Fig 1:- Transmission Section Video Resolution: 1280p x 720p
Camera Sensor: CMOS
 Reception Section Storage: Phone Internal Storage/SD Card Storage
Reception section includes a PC where the image is
processed and abnormalities are detected which is then
uploaded on the server.

IJISRT19DEC582 www.ijisrt.com 568

Volume 4, Issue 12, December – 2019 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
On receiving the images they are greyscale and
number of potholes are detected using canny filter and
prewitt filter as shown in the result.

Fig 3:- Pluto X

 Software Components used

 Python
Python is a general purpose, interpreted programming
language majorly used for web development, software
development, mathematical calculations and such complex Fig 5:- Detection of Potholes
operations. It’s a popular programming language owing to
its easy syntax as well as large scale user community VIII. RESULTS
support, since it is open source.
No. of potholes are detected and displayed on the
Python has a robust standard library, which allows to monitor as shown. This data is then uploaded on
undertake various development, analysis and production ThingSpeak- IoT based web server for plotting graphical
tasks without writing the actual code. With immense support presentation of it. This is useful to detect and identify the
from python libraries, calculations, analysis, etc. are carried number of potholes per unit length of road (100m).
out smoothly.
An observation specimen was also recorded pre and
OpenCV -Python is a Python library designed to solve post monsoon. This corelates number of potholes before and
computer vision problems. OpenCV-Python makes use after rainy season.
of NumPy, which is a highly optimized library for numerical
operations with a MATLAB-Style Syntax. Computer reads
image data as arrays and hence for processing and
manipulating images. NumPy converts data into array
structures and these array structures are used for computer
vision.

VII. OBSERVATION

Pictures of potholes and humps were taken by the

UAV as shown below, and were send to the PC through
wireless communication using ESP12F module, pre-
integrated in Pluto X.

Fig 6:- Pothole detection before Monsoon

Fig 4:- Camera image of road

IJISRT19DEC582 www.ijisrt.com 569

Volume 4, Issue 12, December – 2019 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
[3]. S. Yamuna, S. Raja, Pavithra. M, Ranjitha. M,
Shubashni. R, Sneha. S, “Automatic Pothole Detection
System using Laser,” International Journal for
Research in Applied Science & Engineering
Technology (IJRASET) Volume 5 Issue III, March
2017.
[4]. Gaurav Singh, Vijay M. Purohit, “PORTABLE LOW
COST SOIL HEALTH MONITORING SYSTEM,”
International Journal of Technical Research and
Applications e-ISSN: 2320-8163, www.ijtra.com
Volume 5, Issue 3 (May-June 2017), PP. 125-128.
[5]. Y. Pan, X. Zhang, M. Sun, Q. Zhao, “OBJECT-
BASED AND SUPERVISED DETECTION OF
POTHOLES AND CRACKS FROM THE
PAVEMENT IMAGES ACQUIRED BY UAV,” DOI:
10.5194 September 2017
Fig 7:- Potholes detection after Monsoon
[6]. Akshata Bhat, Pranali Narkar, Divya Shetty, Ditixa
Vyas, “Detection of Potholes using Image Processing
This information can be made public or provided to the
Techniques,” IOSR Journal of Engineering
government authorities for precise pothole location and swift
(IOSRJEN), International Conference on Innovative
repairment of roads.
and Advanced Technologies in Engineering (March-
2018
IX. CONCLUSION
[7]. Seung-Ki Ryu, Taehyeong Kim, and Young-Ro Kim,
“Image-Based Pothole Detection System for ITS
Potholes and humps are dangerous hurdles on the road
Service and Road Management System,” Mathematical
and it should be eradicated as soon as possible. The current
Problems in Engineering Volume 2015.
system includes the use of manual detection by people who
[8]. Ameya Wagh, Zalak Dave, Gaurav Singh, Varun
are willing to contribute for the betterment of the road. Thus,
Dange, Abhay Tambe, Sachin Gengaje, “A low-cost
it is important that manual labor approach is kept to a
postable oscilloscope for educational platforms using a
minimum and switched to an automatic approach instead.
Programmable System on Chip,” 2014 International
Conference on Advances in Communication and
Hence it can be concluded that the number of potholes
Computing Technologies (ICACACT 2014)
before monsoon is less than the number of potholes after
[9]. Aparna, Yukti Bhatia, Rachna Rai, Varun Gupta,
monsoon.
Naveen Aggarwal, Aparna Akula, “Convolutional
neural networks based potholes detection using
The system is used to detect the potholes through an
thermal imaging” Journal of King Saud University -
aerial view which then later processed through a software,
Computer and Information Sciences Available online 5
aids the authority about the abnormality so that they can take
February 2019
the required steps immediately.
[10]. Jia Juang Koh, Timothy Tzen Vun Yap, Hu Ng, Vik
Tor Goh, Hau Lee Tong, Chiung Ching Ho, Thiam
The system dynamically detects potholes in real time
Yong Kuek, “Autonomous Road Potholes Detection
without damaging the cars used for hurdle detections. Thus,
on Video,” August 2018
making the system more feasible, favorable, adaptable and
[11]. Indreswari Suroso, Erwhin Irmawan, “Analysis Of
less expensive and complex.
Aerial Photography With Drone Type Fixed Wing In
Kotabaru, Lampung,” May 2018
ACKNOWLEDGEMENT
[12]. Artis Mednis, Girts Strazdins, Reinholds Zviedris,
Georgijs Kanonirs, “Real Time Pothole Detection
This research was supported by Drona Aviation, a
Using Android Smartphones with Accelerometers,”
drone manufacturing Indian Company, which provided
2011 7th IEEE International Conference on Distributed
insight and expertise that greatly assisted the research.
Computing in Sensor Systems and Workshops
(DCOSS)
REFERENCES
[13]. GOVERNMENT OF INDIA MINISTRY OF ROAD
TRANSPORT & HIGHWAYS TRANSPORT
[1]. Supratik Kumar Saha, Rishav Bhatt, Adit Gavankar,
RESEARCH WING NEW DELHI, “ROAD
Prof. A.P. Singh, “A REVIEW PAPER ON
ACCIDENTS IN INDIA – 2017”, Table (3.4), 15
EXISTING POTHOLE DETECTION METHODS,”
[14]. Gaurav Singh, Siddharth More, Shivprasad Shetty,
International Research Journal of Engineering and
Romit Pednekar, “Implementation of a wireless sensor
Technology (IRJET), Volume: 05 Issue: 12, Dec 2018
node using Programmable SoC and CC2500 RF
[2]. Youngtae Jo and Seungki Ryu, “Pothole Detection
module”, 2014 International Conference on Advances
System Using a Black-box Camera,” Published online
in Communication and Computing Technologies
2015 Nov 19.
(ICACACT 2014)

IJISRT19DEC582 www.ijisrt.com 570

Volume 4, Issue 12, December – 2019 International Journal of Innovative Science and Research Technology
ISSN No:-2456-2165
[15]. Mritunjay Ojha, Shanitamol Sojan, Frezy Roy, Lijo P
Varghese, “An Android Application for School
Information System,” International Journal of
Innovations in Engineering and Technology (IJIET),
2019.
[16]. Suet-Peng Yong, Yoon-Chow Yeong, “Human Object
Detection in Forest with Deep Learning based on
Drone’s Vision,” 4th International Conference on
Computer and Information Sciences (ICCOINS), 2018.

IJISRT19DEC582 www.ijisrt.com 571