Ijaems apr-2016-17 Raspberry PI Based Artificial Vision Assisting System for Blind Persons

International Journal of Advanced Engineering, Management and Science (IJAEMS) [Vol-2, Issue-4, April- 2016]
Infogain Publication (Infogainpublication.com) ISSN : 2454-1311
www.ijaems.com Page | 145
Raspberry PI Based Artificial Vision Assisting
System for Blind Persons
A. Neela Madheswari, R. Dinesh Kumar, R. S. Sabarinathan, M. Manikandan
Department of CSE, Mahendra Engineering College, Namakkal, India
Abstract— The main aim of this paper is to implement a
system that will help blind person. This system is used by
a RASPBERRY PI circuit to provide for the identification
of the objects, the first level localization. It also
incorporates additional components to provide more
refined location and orientation information. The input
process is to capture every object around 10m and it is
convert into the output processing in voice command
which is adopted in Bluetooth headset which is used by
blind people using RASPBERRY PI component.
Keywords— Raspberry PI, artificial vision, Python,
object identification.
I. INTRODUCTION
There are approximately 38 millions of people across the
worldwide mainly in developing countries who are blind
and visually impaired, over 15 million from India. Blind
persons most of the time are withdrawn from the society
because they feel that people and the society are
prejudiced and they may not be welcomed most of the
time [1]. Independent mobility is one of the most pressing
problems facing people who are blind. According to data
published by World Health Organization in 2014, 285
million people are estimated to be visually impaired
worldwide, of whom 39 million are blind and 246 million
have low vision. This means that someone in our world
goes blind in every five seconds [2]. There are many
factors that contribute to the low take-up of electronic
travel aids by blind and visually impaired people.
Technology does not operate in isolation, it must be
considered within the broader context. Users interact with
technology to perform tasks within a social, economic,
political and physical environment.
Due to the development of modern technology, many
different types of navigational aids are now available to
assist the blinds [3], [4], [5], [6]. But almost all the
systems use sensor devices. The objects can be identified
using sensor components. The usage of sensors is
expensive and unsuitable for nowadays. The proposed
work is an attempt to object identification for blind
persons using Raspberry Pi, head phone and camera.
Using this work, the size of the system can be reduced.
There is no necessary for Internet connectivity and the
output through voice makes the process user friendly.
II. SYSTEM ENVIRONMENT
The block diagram of the system is given in Figure 1.
There are three main components namely: Raspberry Pi,
Camera and Headset.
Fig. 1: Block diagram of the proposed system
2.1. Raspberry Pi
The Raspberry Pi is a credit-card sized computer that
plugs into your TV and a keyboard, which can be used for
many of the things that our average desktop does -
spreadsheets, word-processing, games and it also plays
high-definition video. Pi is based on a Broadcom SoC
(System of Chip) with an ARM processor [~700 MHz], a
GPU and 256 to 512 MB RAM. The boot media is an SD
card [which is not included], and the SD card can also be
used for persist data. RAM and processing power are not
nearly close to the power house machines we might have
at home, these Pi’s can be used as a cheap computer for
some basic functions, especially for experiments and
education. The Pi comes in three specifications. It is given
in Table 1. The cost of a Pi is around $35 for a B Model
and is available through many online and physical stores.
In this work, Raspberrry Pi Model B is used. The figure 2
shows the architecture of Raspberry Pi Model B.

International Journal of Advanced Engineering, Management and Science (IJAEMS)
Infogain Publication (Infogainpublication.com
www.ijaems.com
Fig. 2: Raspberry Pi Model B architecture
Table. 1: Raspberry Pi Specifications
Description Model A Model B
Chip
Broadcom BCM2835 (CPU, GPU, DSP,
SDRAM, single USB port)
Processor
700 MHz ARM1176JZF-S core (ARM11
family, ARMv6 instruction set)
RAM 256 MB 512 MB
USB
1 (direct
from
BCM2835
chip)
2 on board
Storage SD card SD card
Voltage
600mA upto
1.2A@ 5v
750mA
upto 1.2A
@ 5v
GPO 26 26
2.2. Camera
Camera Pi is an excellent add-on for Raspberry Pi to take
pictures and record quality videos with the possibility to
apply a considerable range of configurations and effects.
Any type of web camera can be used. For example, a web
camera shown in figure 3 can also be used for this work.
There are some tools helpful for acquiring images and
configuring and obtaining useful information from
images. The usage of OpenCV (Open Computer Vision)
and SimpleCV (Simple Computer Vision) frameworks
that allows simplified usage with Python language.
OpenCV is a specific library for computer vision,
SimpleCV is usable in Python for easy to use and enhance
functionalities of the OpenCV library and image
processing algorithms into higher level ‘bricks’ that
simplify the life of the developer that wishes to create
artificial vision applications without necessary to possess
a deep knowledge of computer vision.
Infogainpublication.com)
Raspberry Pi Model B architecture
Raspberry Pi Specifications
Model B Model B+
Broadcom BCM2835 (CPU, GPU, DSP,
SDRAM, single USB port)
S core (ARM11
family, ARMv6 instruction set)
512 MB
2 on board 4 on board
MicroSD
card
upto 1.2A
600mA upto
1.8A@ 5v
40
on for Raspberry Pi to take
pictures and record quality videos with the possibility to
apply a considerable range of configurations and effects.
Any type of web camera can be used. For example, a web
an also be used for this work.
There are some tools helpful for acquiring images and
configuring and obtaining useful information from
images. The usage of OpenCV (Open Computer Vision)
and SimpleCV (Simple Computer Vision) frameworks
ed usage with Python language.
OpenCV is a specific library for computer vision,
SimpleCV is usable in Python for easy to use and enhance
functionalities of the OpenCV library and image
processing algorithms into higher level ‘bricks’ that
e of the developer that wishes to create
artificial vision applications without necessary to possess
Fig. 3: Web
Using python coding, the steps involved are: i) acquire an
image and save in any of the folders,
particular image, iii) display that image, iv) changes the
original picture into black and white without shades of
gray, v) desired spots for objects are identified and
displayed.
The series of commands to be worked are
follows:
1. apt-get update
It is used to download the package lists from the
repositories and updates them to get information on
the newest version of packages and their
dependencies.
2. apt-get upgrade
It is used to run the updates and upgrades.
3. raspistill –t 5000
This will display the previous window for 5 seconds.
4. apt-get install ipython python
python-numpy python-setuptools python
5. pip install
https://github.com/sightmachine/SimpleCV/zipball/
master
These commands are used to install SimpleCV.
6. pip install svgwrite
This command is used to install the
python module.
7. simplecv
This command is used to verify the running of SimpleCV.
For object identification, the python code written and run
was Listatol.py. It is shown in figure
the objects using Photoshop also for getting clear
identification of images. One of the sample object
identification processes is given in figure
binarization, the picture will be seen in black and white
colours and is shown in figure
Fig. 4: While running Listatol.py file
[Vol-2, Issue-4, April- 2016]
ISSN : 2454-1311
Page | 146
Web Camera
Using python coding, the steps involved are: i) acquire an
image and save in any of the folders, ii) load that
particular image, iii) display that image, iv) changes the
original picture into black and white without shades of
gray, v) desired spots for objects are identified and
The series of commands to be worked are explained as
It is used to download the package lists from the
repositories and updates them to get information on
the newest version of packages and their
It is used to run the updates and upgrades.
s will display the previous window for 5 seconds.
get install ipython python-opencv python-scipy
setuptools python-pip
pip install
https://github.com/sightmachine/SimpleCV/zipball/
These commands are used to install SimpleCV.
This command is used to install the missing svgwrite
This command is used to verify the running of SimpleCV.
For object identification, the python code written and run
was Listatol.py. It is shown in figure 4. We can visualize
the objects using Photoshop also for getting clear
identification of images. One of the sample object
identification processes is given in figure 5. After
binarization, the picture will be seen in black and white
figure 6.
While running Listatol.py file

Infogain Publication (Infogainpublication.com
www.ijaems.com
Fig. 5: Object identification Process
Fig. 6: After binarization
2.3. Head set
Any type of head set can be used for this work. It is used
to receive the audio voice generated by the python coding
after identification of object to user.
III. OBSERVATIONS
The system is connected as per figure 1. We can use
either AC supply or DC supply for Raspberry Pi board.
The head phone is connected to Raspberry Pi using one of
the USB connectors. Any computer system can be
connected to Raspberry Pi using LAN cable through USB
connector in Raspberry Pi. Already an Ethernet card is
attached with the Raspberry Pi. We can make use of any
type of SIM using that Ethernet provision.
First, we have to check the connectivity of
the Raspberry board. Then using the remote system
execution and connection support and testing, system has
to be tested. The entire coding is placed in the SD card
and it is to be inserted into the Raspberry Pi board. Then
the object which is found near the web camera will be
identified.
The sequence of steps involved are: i) ping
IP_address_of_device, ii) perform the remote connection
device using mstsc command, iii) provide the IP address
Infogainpublication.com)
Object identification Process
Any type of head set can be used for this work. It is used
to receive the audio voice generated by the python coding
OBSERVATIONS
The system is connected as per figure 1. We can use
either AC supply or DC supply for Raspberry Pi board.
The head phone is connected to Raspberry Pi using one of
the USB connectors. Any computer system can be
connected to Raspberry Pi using LAN cable through USB
connector in Raspberry Pi. Already an Ethernet card is
attached with the Raspberry Pi. We can make use of any
type of SIM using that Ethernet provision.
First, we have to check the connectivity of the device to
the Raspberry board. Then using the remote system
execution and connection support and testing, system has
to be tested. The entire coding is placed in the SD card
and it is to be inserted into the Raspberry Pi board. Then
found near the web camera will be
The sequence of steps involved are: i) ping
IP_address_of_device, ii) perform the remote connection
device using mstsc command, iii) provide the IP address
for remote desktop connection, iv) provide the
authentication detail to get into the system, v) run the
python coding for obtaining results, vi) Keep the object
near to the camera, vii) Object identified and audio played
and is also the specified object name is also displayed in
the screen.
Using this work, three objects were tested and obtained
the results in a correct manner. Those objects are: i)
mobile, ii) bottle, iii) ball.
IV. CONCLUSION
The technologies are growing day by day. The usage of
technologies for basic needs has to be improved always.
This work is a small contribution for object identification
which will be helpful for blind persons. In the future, we
can extend the work to a many number of objects for
identifications and also the image even captured in poor
light illumination has also to be iden
REFERENCES
[1] Anushree Harsur, Chitra.M, Voice based Navigation
System for Blind people using Ultrasonic Sensor,
International Journal on Recent and Innovation
Trends in Computing and Communication, Vol 3,
Issue 6, June 2015, 4117
[2] Kun Li, Electronic Travel Aids for Blind Guidance,
An Industry Landscape Study, Project Report, 2015
[3] Shachar Maidenbaum, Shlomi Hanassy, Sami
Abboud, Galit Buchs, Daniel
Levy-Tzedek, Amir Amedi, The ‘Eye Cane’, a new
electronic travel aid for the blind: Technology,
behavior & swift learning, Restorative Neurology
and Neuroscience 32 (2014) 813
[4] Manoj Badoni and Sunil Semwal
and Water Pit Indicator using AVR ATmega8 in
Electronic Travel Aid for Blind, International Journal
of Disaster Recovery and Business Continuity, vol 2,
Nov 2011, 1-8.
[5] Kumar A, Patra R, Manjunatha M, Mukhopadhyay J,
An electronic travel aid for navigation of visually
impaired persons, Third IEEE International
Conference on Communication Systems and
Networks (COMSNETS), 2011, 1
[6] Nithya S, Shravani A.S.L, Electronic Eye for
Visually Impaired Persons, International Journal of
Emerging Technology and Advanced Engineering,
Vol 3, Issue 10, Oct 2013, 700
[Vol-2, Issue-4, April- 2016]
ISSN : 2454-1311
Page | 147
for remote desktop connection, iv) provide the
tication detail to get into the system, v) run the
python coding for obtaining results, vi) Keep the object
near to the camera, vii) Object identified and audio played
and is also the specified object name is also displayed in
three objects were tested and obtained
the results in a correct manner. Those objects are: i)
CONCLUSION
The technologies are growing day by day. The usage of
technologies for basic needs has to be improved always.
is a small contribution for object identification
which will be helpful for blind persons. In the future, we
can extend the work to a many number of objects for
identifications and also the image even captured in poor
light illumination has also to be identified better.
REFERENCES
Anushree Harsur, Chitra.M, Voice based Navigation
System for Blind people using Ultrasonic Sensor,
International Journal on Recent and Innovation
Trends in Computing and Communication, Vol 3,
Issue 6, June 2015, 4117-4122.
, Electronic Travel Aids for Blind Guidance,
An Industry Landscape Study, Project Report, 2015.
Shachar Maidenbaum, Shlomi Hanassy, Sami
Abboud, Galit Buchs, Daniel-Robert Chebat, Shelly
Tzedek, Amir Amedi, The ‘Eye Cane’, a new
for the blind: Technology,
behavior & swift learning, Restorative Neurology
and Neuroscience 32 (2014) 813–824.
Manoj Badoni and Sunil Semwal, Discrete Distance
and Water Pit Indicator using AVR ATmega8 in
Electronic Travel Aid for Blind, International Journal
of Disaster Recovery and Business Continuity, vol 2,
Kumar A, Patra R, Manjunatha M, Mukhopadhyay J,
aid for navigation of visually
impaired persons, Third IEEE International
Conference on Communication Systems and
Networks (COMSNETS), 2011, 1-5.
Nithya S, Shravani A.S.L, Electronic Eye for
Visually Impaired Persons, International Journal of
hnology and Advanced Engineering,
Vol 3, Issue 10, Oct 2013, 700-703.

Ijaems apr-2016-17 Raspberry PI Based Artificial Vision Assisting System for Blind Persons

More Related Content

Viewers also liked (20)

Similar to Ijaems apr-2016-17 Raspberry PI Based Artificial Vision Assisting System for Blind Persons (20)

Recently uploaded (20)

Ijaems apr-2016-17 Raspberry PI Based Artificial Vision Assisting System for Blind Persons