2018 3rd International Conference for Convergence in Technology (I2CT)

The Gateway Hotel, XION Complex, Wakad Road, Pune, India. Apr 06-08, 2018

Automatic Video processing based on IoT using

Raspberry Pi

Vaishnavi Mande Mandar Lakhe

Electronics and Telecommunication Engineering Electronics and Communication Engineering
Yeshwantrao Chavan College of Engineering Jigyasa Research and Development centre
Nagpur, India Nagpur, India
[email protected] [email protected]

Abstract— In this era of modernization, use of Internet has this model to bring forth a more convenient way of automatic
increased to a huge extent. Automation brought about by video processing.
technology has saved human efforts to a large extent. The
Internet of Things has revolutionized the way of communication.
Communication is a powerful tool to exchange information and II. METHODOLOGY
ideas. The evolution in science and technology which has taken
place in the last decade is commendable. Video communication
A Capturing of Image
has quickly become a business standard in many industries.
Video communication aids in Telecommuting, humanizing the In this process capturing of video, images takes place. This
conversation, bridging the communication gap and much more. paper is all about the automatic video processing using
Video processing is an important process of video Raspberry Pi. So the video of a particular background is
communication. Many parameters like video quality, data losses, captured
cost effectivity of model, relative humidity, fog play an important
role. Video communication allows businesses to commute more
effectively. Hence, a digital system for automatic video processing
based on the concept of Internet of Things using Raspberry Pi’s B. Proposed Architecture
has been proposed.
Below diagram depicts the proposed architecture of our
Index Terms— Raspberry Pi, image processing,video
conferencing, video processing, OpenCV, Internet Of Things, model. This model basically consists of four stages. Each
GUI components. stage has been explained. This model processes videos,
images as well as texts.
)

I. INTRODUCTION
Today, the Internet is the most powerful tool in the world.
Internet has made our lives better. The Internet has drastically
improved every aspect of human life. The foremost target of
Internet has always been communication. A revolutionary
technology, called as the Internet of Things (IoT) came out of
Internet. The Internet of Things is the biggest wave in the
development of Internet. There are many real time
applications of IoT like smart cities, smart supply chain,
precision agriculture, healthcare and many more.
Using the concept of Internet of Things, a model has been
proposed for automatic video processing using Raspberry Pi.
The tremendous bandwidth required by video communication
makes it one of the most challenging application. Video
processing is one of the most important step of Video
communication. A number of problems have been tackled with
the help of this model. Raspberry Pi, which acts as a mini
computer , has an excellent video quality. It has been used in
Fig.1. Proposed achitecture


978-1-5386-4273-3/18/$31.00 ©2018 IEEE 1

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on July 28,2022 at 06:39:43 UTC from IEEE Xplore. Restrictions apply.
C. Power Supply Windows, Linux and Android are also supported. Many
In this model we have proposed to use +5.1V micro USB modules can be inbuilt in Python using C and C++. In this
supply. The current (milliAmperes) required to run a Raspberry proposed model, imported libraries of C++ have been binded
Pi is generally dependent on what it is connected to. The with Python. It has been used for video, image and text
maximum power a Raspberry Pi can use is 1 Ampere. The processing. But it has most widely been used for video
power requirements of a Raspberry Pi increases as more processing. Sensors will fetch all the data regarding text, image
interfaces are connected to it. or video processing. All the settings for resolution, rotating
angles, etc are done by using many algorithms available in
Open CV. Raspberry Pi’s will receive data via sensors.
D. GUI Components
GUI stands for Graphic User Interface. Well-designed
Graphical User Interfaces frees the users from learning
complex language commands. It is a way to communicate
what one wants a computer application to perform using
graphical symbol. In this proposed model, a GUI screen has
been placed upon both the Raspberry Pi’s. In this proposed
model, a screen will be placed upon both the Raspberry Pi’s.
There will mainly be three buttons in the GUI. The buttons
will be used for video, text and image displaying purposes.
Along with these three buttons, two more additional
toolboxes would be present as well. One would be for
textboxes, that is entering text and the other would be for
taking axis of images into consideration. The images can be
in any format like JPEG, MPEG, GIF.

VideoVideo Fig.3. Video conferencing python code

ImImagext

Textge

Fig.2.Raspberry Pi GUI screen

E. OpenCV communication flow

Open CV stands for Open Source Computer Vision Fig.4. Image processing python code
Library. It is an open source computer vision and machine
learning software library. In this model, this concept has been
used for video processing. It consists of C++, C, Python,
MATLAB, and Java. Operating Systems like Mac OS,


2

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on July 28,2022 at 06:39:43 UTC from IEEE Xplore. Restrictions apply.
 recognize faces, classify human action in videos, track camera
movements, extract 3D movements of objects, find look alike
images from database, etc. OpenCV was started to provide a
common platform for computer vision applications. It supports
C++, C, Python, Java, MATLAB languages. Operating
Systems such as Mac OS, Windows, Linux, Android. OpenCV
focuses mostly on real time vision applications. OpenCV-
Python is a library of Python bindings designed to solve
computer vision problems. Python is one of the simplest
general purpose programming language. It is very famous
because it has good code readability and quite simple to use as
well. But Python can be easily extended with C/C++, using
which computationally intensive codes can be written in
C/C++. OpenCV-Python is a Python wrapper for the original
OpenCV C++ implementation. OpenCV-Python makes use of
Numpy library which is used for numerical operations in a
Matlab style code language.

Fig.5. Text processing python code

IV. RASPBERRY PI DESCRIPTION

The Raspberry Pi Foundation introduced the first Raspberry

Pi in 2012. The main idea behind inventing Raspberry Pi was
to to promote teaching in schools and colleges. Raspberry Pi is
like a small credit card sized computer which uses a keyboard
and mouse. It has got various functionalities like browsing
F. Communication between Raspberry Pi’s
internet, playing high definition videos, making spreadsheets,
word processing and playing games. It is slower than a
Communication between both the Raspberry Pi’s can take modern desktop or a laptop. But it acts as a complete Linux
place in three ways: Operating System computer and has the ability to perform
1.Video Processing: In this step, pixels in video are deciphered. many high end functionalities. It consumes very less power as
Accordingly, brightness and sharpness of image is set. well. The Raspberry Pi computer is essentially a wireless
Resolution settings are taken into consideration as well. Internet capable system-on-a-chip (SoC) with 1 GB RAM,
2. Image Processing: It will operate in the same way as Video connection ports, a Micro SD card slot, camera and display
Processing. The Raspberry Pi will receive frames for interfaces and an audio/video jack. There is numerous amount
processing the images at a rate of 30frames per second. of advantage in using Raspberry Pi. Because of the small size,
3. Text Processing: In this, the textual data received is it works as a normal computer which can handle online traffic.
processed. The raspberry pi finds usage in various application. It can be
used for like Media streaming, Computer Tabs, Home
automation devices, Monitoring robots, Cosmology, and also
G. Action Executed
in raspberry pi based projects.
In this step, the video will be played on the camera of slave
as captured by the Master Raspberry Pi using the OpenCV
code. In the same way, images as well as texts are also
transferred

III. OPENCV LIBRARY

OpenCV stands for Open Source Computer Vision Library.

It is an open source computer vision and machine learning
software library. OpenCV had started way back from 1999
and the first model was launched in 2000. This library has
more than 2500 optimized algorithms, which includes a
comprehensive set of machine learning algorithms as well as
Fig.6. Raspberry Pi circuit board
for computer vision. These algorithms can be used to detect an


3

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on July 28,2022 at 06:39:43 UTC from IEEE Xplore. Restrictions apply.
. V. INTERNET OF THINGS based. As soon as the video button on the GUI of Raspberry Pi
is played, the camera of Master Raspberry Pi starts recording
the video. The Master Raspberry Pi will configure the IP of
slave one and automatically the video processing of the video
will be done using the OpenCV code which is stored on the
video button of Master Raspberry Pi. This process is called as
video processing. As soon as video processing is done, the
video will be played in the Slave Raspberry Pi’s camera after
pressing the video button on the GUI of Slave. It will play the
video captured by Master .In the same, if Slave Raspberry Pi
is operated first, it will configure the IP of Master Raspberry
Pi and do the same process as done by Master one initially.
This leads to both way communication between Raspberry
Pi’s. A cluster of Raspberry Pi’s can also be made depending
on the need of video conferencing devices. If a third
Raspberry Pi is added to this model, it can configure the IP
address of Master or Slave whichever is needed and hence
plays the video. In video processing, the frames are processed
at the rate of 30 frames per second. In the same way, two more
additional buttons are present on the GUI for text and image
transfer. They too operate in the same way as Video transfer.
After pressing the text button on GUI, the text will be
transferred to the camera using the Text code on the text
button. The image transferring takes place in the same way.

Fig.7. Internet of Things

The concept Internet of Things was not known until the late
90s.The Internet of Things concept has been in process for
decades. The first appliance made using Internet Of Things
was a coke machine in the year 1980 .The programmers could
make out the whole functionality of a machine including
getting continuous updates and predicting the results as well.
Internet of Things is the huge network of devices connected to
Internet consisting of mobile phones, electrical appliances,
and engines. Any device with a sensor on it is also used. These
things collect and exchange data. Each thing is uniquely
indentifiable through its embedded computing system. These
devices collect data with the help of existing technologies and
then using automation the data flows between other devices.
But security and privacy are the biggest challenges for this
technology as a huge risk of personal information is involved.
This technology also suffers from unstable platforms and not
so good technology standards. It is mainly an ecosystem
dependent technology.

VI. WORKING

In this model, we have used two Raspberry Pi’s. The

Raspberry Pi’s have been marked as Master and Slave
depending on who we operate first. Using the concept of
Internet of Things, both of them have their respective IP
address. If we use first Raspberry, then it will become the
Master and the second Raspberry Pi will become the Slave. In Fig.8. Hardware Architecture
the same way it follows vice versa. The whole setup is IOT


4

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on July 28,2022 at 06:39:43 UTC from IEEE Xplore. Restrictions apply.
 VII. SYNTHESIZED RESULTS
Simulation is the process of verifying the various
functionalities of a model using key functions and operations.
To our knowledge it is the first time that automatic video
processing using Raspberry Pi and OpenCV library based on
Internet of Things (IOT) has been used. The simulation has
been carried out in Python language.

Fig.11. Current Vs Environmental conditions plotted graph

VIII. CONCLUSION

Using this proposed model, day to day problems faced by

people while video conferencing have been dealt. Especially,
the use of humidity/fog sensors ensures no problems are faced
in the regions having high relative humidity or foggy regions.
Since this model is very cost effective ,as the requirement
increases, more Raspberry Pi’s can be joined together to form
Fig.9. Video conferencing on multiple cameras a cluster. There are almost negligible data losses. This model
uses OpenCV for performing many applications. With the
help of many algorithms from OpenCV, numerous
applications can be added as the need arises. The video quality
is excellent as Raspberry Pi has a full HD 1080 Media Player.
From the data collected in relation to Current versus Cost effectivity is very high as many interfaces can be
Environmental conditions, it has been found that as the current connected to Raspberry Pi to act for it as a normal computer.
increases, so does the environmental conditions like fog and The Graphic capabilities are even better than the Apple
humidity sensors increases. The increase is not directly Products. Hence, this model to a great extent solves all the
proportional. The current measured is in milliAmperes. problems faced while video conferencing.

REFERENCES
[1] J. Rahman, Motion detection for video surveillance,
http://www.divportal.org/smash/get2:518464/FULLTEXT01.pdf
[2] R. Szeiliski, Computer vision : Algorithms and Application.
http://szeliski.org/Book/.
[3] Raspberry Pi Foundation, “Quick Start Guide”.guide - v1.1.pdf
[4] XBMC Foundation, “Official XBMC software”,”
http://xbmc.org/about/xbmc-software”.
[5] Performance, Management and monitoring of 68 node Raspberry
Pi 3 Education cluster: Big Orange Bramble(BOB), The
University of Tenesse, knoxville, USA.
[6] Implementation of Image processing on Raspberry Pi, K.S.
Shilpashree, Lokeha.H, International Journal of Advanced
Fig.10. Proposed attempted model Research in Computer and Communication ngineering.
[7] Industrial Automation using IoT with Raspberry Pi, H.K.
Merchant. D.D.Shine, International Journal of Computer
Applications(0975-8887), volume 168 – No.1, June 2017.


5

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on July 28,2022 at 06:39:43 UTC from IEEE Xplore. Restrictions apply.
[8] Autonomous Object Detection and Tracking using Raspberry Pi,
Sapna Jana, Shubhangi Borkar, IJESC.

Authorized licensed use limited to: INDIAN INSTITUTE OF TECHNOLOGY DELHI. Downloaded on July 28,2022 at 06:39:43 UTC from IEEE Xplore. Restrictions apply.