Proceedings of the IEEE 2017 International Conference on Computing Methodologies and Communication

(ICCMC)

Design of IoT Based Intelligent Parking System

Using Image Processing Algorithms
Sagar Rane Aman Dubey, Tejisman Parida
Assistant Professor, Computer Engineering Students, Computer Engineering
Army Institute of Technology, Pune Army Institute of Technology, Pune

Abstract— Advancement in industrialization leaves the parking person who wants to park his/her car, then they should find
management system outdated. Traffic condition in foreign parking location which is time consuming. In that situation,
countries are homogeneous whereas India’s traffic condition is improper parking lot management can trouble them a lot. One
heterogeneous which makes real-time management at parking lot situation can be that the person will not find any position to
difficult. Traffic congestion at parking lot has become often, that
park vehicle, whereas parking is available. So, it is very
people cannot even find the place to park their vehicle especially
in large companies where first come first serve method is used for important to design an intelligent parking system to control and
parking vehicle. In nutshell, parking condition needs to be manage vehicles at parking efficiently [4-5]. The most
improved by introducing Internet of Things (IoT) to parking lots. prevalent reason of traffic congestion at parking is an
With the help of Microcontrollers and Image processing, inefficient controlling which affects the traffic flow. For
tradition parking lots can be converted into smart one. The example, assume a situation where parking is available in
proposed system will also help company to enter only registered parking lot and people cannot find that location because the
employees and to manage limited parking place, efficiently. parking lot is big, then this is the waste of available resources
Image processing algorithms, such as OCR (Optical Character and is inefficient. This happens because early coming people
Recognition) are used to check valid employee. After image
park their vehicle where ever they want to. By considering the
processing algorithm, number would be extracted out from
number plate and sent to server, for employee verification. In the above example if proper count of vehicle is maintained and
similar fashion, server will also receive exact traffic density count monitored then people need not to unnecessarily find the
at parking lot and other statistics from the microcontroller. parking position, and this might solve the problem.
Android App can be used by the user to extract the details from
server at any time on demand find the proper location for
Techniques proposed to control the traffic signal by using
parking. This technology will help the companies with large image processing algorithms [2] (in OpenCV) are: OCR Scan
number of employees to park their vehicle without wastage of which stands for Optical Character Recognition. Through
time. This system also alert authorities when less than 10% space character recognition, number plates can be read. Camera will
is available and have reserved parking for women, physically feed the car front image containing number plate. This image
handicapped person, and pregnant women. For traffic will be preprocesses and converted to Greyscale Scene and
management at parking lots, a prototype has been designed and Image thresh Scene. In this image, possible characters will be
tested. Main objective is to come up with a solution for found out by making contours of the image. Now the image is
traditional parking lots to save time and money of user. converted to text and this text can be given to server using
Index Terms— Traffic Analysis, Image Processing, Intelligent
beagle bone black. Server will check the number plate and will
controlling of traffic, Broadcasting App conclude that he/she is a valid employee or not. This algorithm
needs to be trained for accurate results, this training takes some
time. By this system, verification of vehicle can be done very
I. INTRODUCTION quickly which saves the time of employee.
Traffic is one of the banes of urban life in India. Before entering the campus of the company, the user will get
While in developed countries, even the richest of the rich use to know that parking position is available or not. If it is
public transport, in India there is a class divide when it comes available, the he will be notified with the floor/area where
to using public transport. This has led to an exorbitant increase parking is available. If in case it is no available, then he will be
in the number of private vehicles leading to an increasing notified before entering the company campus so that the
number of traffic snarls [1]. One of the major problem due to employee can find parking outside the campus. For user to get
increase of vehicles the problems with parking lots. Vehicle the real-time updates, a region of influence is set around the
management at the parking lots have become an important campus using longitude and. When user enters into the region
aspect for the best utilization of existing parking area capacity. then he will automatically get the real-time updates. Along
Wastage of time and fuel at the parking area due to increase in with that user will be provided with the position where parking
vehicle count have become a major issue for big company like will take less time. The paper is organized as follows: section II
Infosys, Accenture etc who are making multilevel parking lot explains the design of the system. Section III discusses the
facility for their employees. It is important to manage the working of the system. Section IV concludes the paper. Finally,
vehicular flow efficiently for the best utilization of existing section V gives future scope of the project followed by the
parking lot capacity. Due to parking lot problem at peak hours, acknowledgement and key references used in the research
human routine also got disturbed. For example, if there is a work.

978-1-5090-4890-8/17/$31.00 ©2017 IEEE 1049

 Proceedings of the IEEE 2017 International Conference on Computing Methodologies and Communication
(ICCMC)

II. SYSTEM DESIGN

FIG. 1. BLOCK DIAGRAM FOR PROPOSED MODEL

The work is divided into 4 parts. First part is the

implementation of image-processing algorithm, where image is
fetched from the camera. Microcontroller will do the
processing part. The second part is to compare the number
plate with the database for valid employee. The third part is to
send the real-time count of each floor/area to the server, so that
authorities and employees can be informed when space
available is less than 10% as well as separate space for women
and physically handicapped can be left out. Finally, an App
will be provided to user as an interface, to get the updates.
Below is the illustration of prototype and the algorithm used.
A. Real Time Image Processing
A system having camera attached with Microcontroller (Beagle
Bone Black / Raspberry Pi) is installed on the entrance.
Microcontroller will be attached to a camera which is
responsible for applying the algorithm [10]. Microcontroller
can perform the OCR Scan at a quick rate and send the needful
data to the server, for validation of the employee. Camera gives
the front image of vehicle to the Microcontroller whenever
vehicle appears according to the algorithm. Applying image Fig. 2. (b) Image with Grey Scale
processing algorithm will give the needful data from the raw
image. This data will then be send to the server for further
processing. We can use OpenCV to do these operations. It is an
open source tool in which we have implemented our prototype.
OpenCV (Open Computer Vision) is used for image
processing. Computer vision is a field of informatics, which
teaches computers to see. It is a way computer gather and
interpret visual information from the surrounding environment
[1].

978-1-5090-4890-8/17/$31.00 ©2017 IEEE 1050

 Proceedings of the IEEE 2017 International Conference on Computing Methodologies and Communication
(ICCMC)

Fig. 2. (c) Finding all possible Characters in Scene In Fig. 2, OCR (Optical Character Recognition) Algorithm is
used, which can be made more accurate by training its
classifier. This classifier is used to make the algorithm more
perfect. According to this Algorithm, the actual scene image is
converted to grey scaled, on which algorithm is applied to filter
out the characters present in the image. In this contours from
vector of possible characters in the scene is found out. At last
matching characters in the scene is extracted out, which gives
the final character present on the number plate.

Fig. 2. (d) Finding contours in vector of possible characters

Fig. 3. Optical Character Recognition Algorithm (OpenCV)

B. Comparing and finding valid employee

Through the OCR Scan car number can be extracted out form
the number plate. Now this number is sent to the server for
Fig. 2. (e) Finding vector of matching Characters in Scene further process. In the server side, company’s database is
present and every employee is registered with their vehicle
number. This database is created according to the need of the
server with the help of Big Data and Hadoop concepts. Map-
reduce algorithm is applied for filtering out the data from the
garbage of all the data present, this is known as Data Mining.

C. Sending Data to the Server

Server-side script receives real time updates and calculated
statistics from the client [9]. The data can be further used for
security purpose also. Apart from saving time in validation
process, the data will also give an extra advantage to the user
for choosing their parking position and to find the nearest
parking in short time. There is a wireless link between
Fig. 2. (f) Characters found in the Scene microcontroller and the central server. All the processed data
will be sent from the microcontroller to the central server and
the central server will make it available on the mobile app.
Database contains the data present on central server, which can
be used at a later stage to extract certain important statistics
that can be used to improve the system further.

Fig. 2. OCR Scan Algorithm

978-1-5090-4890-8/17/$31.00 ©2017 IEEE 1051

 Proceedings of the IEEE 2017 International Conference on Computing Methodologies and Communication
(ICCMC)

D.Extracting Data from Server through an App

IPC (Inter Process Communication) Connection is used to send
the data from the sender to the receiver. Socket connection is
used for the purpose to store data in the database.
Microcontroller will send all the data to the server wirelessly.
Data from different floors in case of multi-level parking and of
different areas in case of ground level parking is updates in the
server and the App also will simultaneously get updated.
III. WORKING AND EXPERIMENTATION
The proposed idea is implemented using Microcontroller, i.e.
Beagle Bone Black using OpenCV. The Microcontroller will
be in picture at the time of vehicle entry and to send the real-
time data to the server. In hardware part, a Camera is attached
to Beagle Bone Black which is connected to the server through
an unguided medium. Camera will click the image of the
vehicle front, which contains number plate. Beagle Bone will
receive the image from camera and simultaneously process the
image to give useful. From image, one can extract out details
like vehicle registered number. Management of traffic at
Fig. 4. (c) Personal parking page (d) Car parking demo
parking and its coordination is possible because after
maintaining count of vehicle entering per floor. Counters at the In Fig. 4(c) user can have a real time look at the floor-wise
entrance can be changed according to the cars getting in and parking lot. The cars with red are booked and remaining
out of the parking lot. These counters are managed and positions are vacant. If the user clicks on any of the position,
coordinated in such a way that waiting time of vehicles will be then it will give an option to book that location. This facility
reduced. According to the counters, data at server end will be will be given to the user only after the security validation
changed. This will maintain a separate reserved space for check. There is a concept called “REGION OF INFLUENCE”
women, pregnant women and physically handicapped person. which is drawn around the campus, so that after entering the
Along with that this will also check if the parking space region of influence the employee/person can be notified
available is less than 10% or not, if it is so then it will update whether the space is available on the parking lot or not.
the server, so that employees will get to know before entering
the campus.
Other area of this project is related to server, databases, and
Android App. Using IPC (Inter Process Communication)
protocol or socket connection the data is given to the server by
microcontroller. User can get the parking updates of the
parking lot at real-time.
In the App, user can create his personal account for saving any
favorite parking position and for validation as shown in
Fig.4(a). User can see the parking position available as shown
in Fig.4(c) and select most suitable position from the available
parking list.

Fig. 5. Region of Influence

The ‘REGION OF INFLUENCE’ is shown by the

encircled area in Fig.5. When the user enters the influence
region, a notification will pop up into the user’s mobile which
shows the real-time updates and other statistics of the
upcoming traffic light. Along with that, most suitable and less
time taking parking position will be given to the user. This
region is set according to user’s requirement, so that user will
have the option to either park into the campus or to choose park
outside, if parking not available inside campus. Suitable
position will be decided by taking all the other parking areas
Fig. 4. (a) Login & Sign Up (b) Side Bar Panel

978-1-5090-4890-8/17/$31.00 ©2017 IEEE 1052

 Proceedings of the IEEE 2017 International Conference on Computing Methodologies and Communication
(ICCMC)

into consideration coming to user’s path. This GPS system will [2] Vikramaditya Dangi, Amol Parab, Kshitij Pawar & S.S Rathod, “Image
Processing Based Intelligent Traffic Controller”, Undergraduate
be established by using longitude and latitude of the campus. Academic Research Journal (UARJ), Vol.1, Issue 1, 2012.
By this way user can directly interact with the parking lot [3] Papageorgiou M., Diakaki C., Dinopoulou V., Kotsialos, “A Review of
before reaching the campus. road traffic control strategies”, Proceedings of IEEE, Vol. 91, Issue 12,
pp. 2043-2067, November 2004.
[4] Georgios Vigos, Markos Papageorgioua, Yibing Wangb, “Real-time
IV. CONCLUSION estimation of vehicle-count within signalized links”, Journal of
Transportation Research Part C: Emerging Technologies, Volume 16,
Managing parking lots is a crucial challenge for the society Issue 1, pp.18–35, February 2008.
and IoT requires to tackle that problem efficiently. By this [5] Michael W. Szeto and Denos C. Gazis, “Application of Kalman
research work, Image processing is demonstrated as an Filtering to the Surveillance and Control of Traffic Systems”, Journal of
efficient technique to validate that whether a person is an Transportation Science, vol. 6 pp. 4419-439 , November 1972.
employee or not. It is shown that the proposed technique can [6] Celil Ozkurt and Fatih Camci, “Automatic Traffic Density Estimation
and Vehicle Classification for Traffic Survillance Systems Using Neural
reduce the problems at parking lots. A microcontroller will Networks”, Journal of Mathematical and Computational Applications,
facilitate a real-time analysis and optimization of vehicles. The Vol. 14, No. 3, pp. 187-196, 2009.
proposed algorithm ensures that the average waiting time of [7] D. Beymer, P. McLauchlan, B. coifman and J. Malik, "A real-time
vehicles at the entrance and to find proper parking will under computer vision system for measuring traffic parameters", IEEE conf.on
any circumstance be lesser than the waiting time required while computer vision and pattern Recognition, (1997), pp. 495-501.
using the currently existing parking systems. By this way, real- [8] Pratishtha Gupta, G.N Purohit, Sweta Pandey, “Traffic Load
Computation for Real Time Traffic Signal Control”, International
time scenarios can be created for user even if user is not Journal of Engineering and Advanced Technology, Vol. 2, Issue 4, April
present at the place. This system is capable of handling 2013.
multiple tasks and is also capable of handling peak hour [9] B. Jin, D. Zou and Y. Gan, "Research and design of traffic detection
situations. Since ‘time is money’, the more time you save, based on GPS", IEEE conf. on advanced computer control (ICACC),
more money you can earn. User’s time can be saved if user (2010) March.
chooses the position given by App. Smart parking system [10] “Beagle Bone Black System Reference Manual”, Revision A5.2April
11, 2013. (http://www.adafruit.com).
implementation saves fuel since user do not have to find the
[11] Pratishtha Gupta, G.N Purohit, Sweta Pandey, “Traffic Load
parking place for vehicle. Sensors are fragile in nature and Computation for Real Time Traffic Signal Control”, International
needs maintenance which makes algorithmic approach more Journal of Engineering and Advanced Technology, Vol. 2, Issue 4, April
suitable for this project. This system will alert authorities in 2013.
case of unidentified vehicles and when space available <10%. [12] B. Jin, D. Zou and Y. Gan, "Research and design of traffic detection
based on GPS", IEEE conf. on advanced computer control (ICACC),
(2010) March.
V. FUTURE SCOPE [13] Georgios Vigos, Markos Papageorgioua, Yibing Wangb, “Real-time
estimation of vehicle-count within signalized links”, Journal of
Transportation Research Part C: Emerging Technologies, Volume 16,
Community based service can also be added to the App, if user Issue 1, pp.18–35, February 2008.
increases so that appropriate parking position can be identified [14] R.J. Oentaryo and M. Pasquier, "Self-Trained Automated Parking
System", International Conference onControl, Automation, Robotics and
even outside the campus. Employees can also share the road Vision, Vol. 14, No. 4, pp.1005, Dec. 2004.
condition on which they are travelling, so that others can avoid [15] L. Hawkins, "The Skinny on Self-Parking", The Wall Street Journal,
that road if that route is busy. Accordingly, the App can be Nov. 2012.
made more efficient. This prototype can be used in mall and [16] R.Roy, "7 Next-Gen Driving Technologies, Coming Soon to BMW",
other places where parking is the major issue. By training & Popular Mechanics, Nov. 2010.
learning, an algorithm can be designed to get the favorite
parking spot of any person. Moreover, this system can also be
used by government, local bodies and police authorities for
security purpose or for any estimation. By sending real-time
status of parking slots, user would be notified before reaching
to the Infosys campus.

VI. ACKNOWLEDGMENT
Financial support and research facilities provided by Army
Institute of Technology, Pune are acknowledged. Authors want
to give special thanks to HoD Computer Department for their
support.

VII. REFERENCES
[1] Sabya sanchi kanojia, “Real –time Traffic light control and Congestion
avoidance system”, International Journal of Engineering Research and
Applications (IJERA), pp.925-929, Vol. 2, Issue 2, Mar-Apr 2012.

978-1-5090-4890-8/17/$31.00 ©2017 IEEE 1053