Recent Trends in Intensive Computing 917

M. Rajesh et al. (Eds.)

© 2021 The authors and IOS Press.
This article is published online with Open Access by IOS Press and distributed under the terms
of the Creative Commons Attribution Non-Commercial License 4.0 (CC BY-NC 4.0).
doi:10.3233/APC210303

Smart Reader Glass for Blind and Visually

Impaired People
S. ANBARASIa 1, S. KRISHNAVENIa, R. ARUNAb
K. KARPAGASARAVANAKUMARc
a
Associate Professor, P.S.R. Engineering College, Sivakasi, TN, India
b
Assistant Professor, P.S.R. Engineering College, Sivakasi, TN, India
c
UG Student, P.S.R. Engineering College, Sivakasi, TN, India

Abstract. Visually impaired people fail to read the text with existing technology. The
proposed project targeted to design a spectacle with a camera by which the blind
visually impaired people can read whatever they want to read based on contemporary
OCR (optical character recognition) technique and text-to-speech (TTS) engines. This
proposed smart reader will read any kind of documents like books, magazines and
mobiles. People can access this novel technology with blindness and limited vision.
The earlier version of the proposed project was developed successfully with mobile
reader which had certain drawbacks such as high cost due to the need of android
mobile, not user friendly and improper focusing. To overcome these disadvantages, a
spectacle type reader with camera is proposed in this project, which will be cost
effective and more efficient.

Keywords: Micro camera, Python, Raspberry pi, Smart reader, Text-to-speech

1.Introduction

Vision is the beautiful gift that God has given to all living creature. Whereas, the
people with blindness and deprived of vision can match with the person with normal
vision by gathering information using their hearing capability only. Most of the visually
impaired people use Braille system for reading documents and books [1]. However, the
Braille system has many disadvantages as follows [2]
• Converting all the materials in to Braille is very expensive
• The Braille materials can’t be read by people with normal vision
• The error in Braille system can’t be corrected easily
• Reading document and books requires more practice and time and this system is not
applicable for all books, documents and many articles.
• It would be very hard to make every visually challenged person to adopt Braille
method.
Even though lots of tools are available in online to convert text to voice there are
certain limitations such as they are not wearable and user friendly, not in compact size
and couldn’t be used commonly to read all materials like books, computers and
mobiles. To overcome the above-mentioned problems our project aimed to design
assistive smart glasses in wearable design format for the blind and visually impaired
people. This smart

1
S. ANBARASI, Associate Professor, P.S.R. Engineering College, Sivakasi, TN, India,
Email: [email protected]
91 S. Anbarasi et al. / Smart Reader Glass for Blind and Visually Impaired People

spectacle with micro camera setup proposed in this project is designed to support
reading printable version of any books, documents, mobile texts by converting text to
audio, which can be heard by microphones or speakers. This portable and economical
smart spectacle is programmed with raspberry pi module and the image processing
technique helps in recognizing and extracting the text from the image [4, 5]. Finally, the
extracted text is converted into speech and can be heard by blind and visually impaired
people. The final hardware model is tested with two test samples, the first one is with
book page and another one is mobile document. The designed smart glass converted
both test samples into a right audio format. This project is very affordable to all
category of people and will be more useful.

2.Objective

In recent days, various learning tools are being developed for blind and visually
impaired people using artificial intelligent techniques [3-5]. However, most of them are
not cost effective. In addition, each reading aids can be used to read a specific
application. The reading assist used in Personal Computer (PC) is not suitable to read
texts in book [6]. Furthermore, not all the reading aids are user friendly, affordable and
compact size. Though various types of reading assistants are present they could not be
used to read multiple texts which is presented in the printed materials or books,
computer screens, and newspaper and mobile screens. In addition, the existing
techniques are not configured for different fonts, colors, scales and multiple
orientations. Hence, to overcome the above-mentioned disadvantages our project aimed
to develop smart reading spectacles for blind and visually impaired people. In our
project, the micro camera is fitted on the reader’s spectacle, which captured the text
image to be read and converted this text image to text again and further it is converted
into audio using text to speech engine. The main objective of this project is to satisfy
the needs of visually impaired people in reading and browsing articles in documents,
computers or in mobiles. This spectacle can also be suggested for normal people during
their travelling time. Since, people getting distraction and tired of reading large number
of pages during travel time. This proposed project can be offered in compact size and
affordable prize due to its simple design.

3.Description of Smart Reading Model

The block diagram representation of Raspberry pi designed smart glass model is shown
in Figure 1. The text image, which has to be converted into audio, is placed under the
micro camera located in spectacle type smart reader. A separate portable rechargeable
power bank arrangement is provided for raspberry pi and camera setup. In raspberry, pi
an assistive system offered text localization algorithm with some accuracy. On startup,
the system verifies its features, matching, alignment, tracking, motion analysis,
deformation estimation and 3D reconstruction. Finally, the good quality of image is
then converted into text and with the help of text-to-speech (TTS) engines; the audio
can be heard through headphones.
 S. Anbarasi et al. / Smart Reader Glass for Blind and Visually Impaired 9

Figure 1. Block diagram representation of smart reader glass

4.Improving Quality with Image Processing

The main challenges faced in computational reading analysis with smart mechanism are
different categories of materials, different colors and different backgrounds. Hence,
appropriate approaches might be used for processing the images to remove noise,
correcting edges and illuminations. In this proposed project the Raspberry Pi embedded
with integrated peripheral devices such as Bluetooth and USB, which can be used for
wireless transformation of data output.
The digital image processing allows to extract clear and required information from
the enhanced image is also used in this proposed project in which the image is
converted to an array of small integers, called pixels, representing a physical quality
such as scene radiance, stored all devices and their connectivity is correct. Further,
computer and other digital hardware modules carried out the segmentation and
extraction process.

5.Working Methodology

The capture button is placed on the spectacle handle. When it is clicked, the document
has to be read will be captured with the camera which is linked to ARM microcontroller
via USB. The selected image is further processed with Optical Character Recognition
(OCR) technology in which the internal process is carried out and the labels are
separated using CV library [7]. Hereby the scanned image is converted in to printed
text. In this proposed project, the TESSERACT library is suggested which is used to
convert text-to-speech. Finally, the printed text is converted into audio using text to
speech engine and it will be pronounced. This voice can be heard through the
microphones [8]. The image to text conversion differs for each document. For written
messages, some special methods such as machine learning might be used for clear
transformation of image to printed text. However, this project mainly aimed to convert
documents in printed textbooks, computers and mobiles and the machine learning
technology is suggested in near future while using to read hand written materials. The
circuit diagram for the proposed smart reader glass was drawn using Software and
depicted in
Figure 2.
92 S. Anbarasi et al. / Smart Reader Glass for Blind and Visually Impaired People

Figure 2. Circuit Design of Smart Reader Glass

6.Flow of Assessment

The methodology of proposed smart reader glass for blind and visually impaired is
described well in section 5. As the image processing plays a vital role in this project it
is essential to describe the process flow for ease of understanding of the readers. clearly
explains the concepts of conversion of the scanned image into an audible text.

6.1 Image Capturing/ Scanning:

A micro camera is fitted on the glass of the spectacle, which is used to capture the
documents to be read after the captured button is clicked.
6.2 Pre-Assessment:
The main processes in Pre-Assessment are skew correction, linearization and noise
removal. The primary step in pre assessment is Skew detection, which is applied to the
scanned text image to convert it into digital format. The second step in pre assessment
is linearization, which supports to linearize the opto-electronical conversion function
(OECF) with three methods such as Full Image Range, Modulation Transfer Function
(MTF) or Full Data Range. The final step in pre-assessment is noise removal in which
the undesirable by-product such as random variation of brightness or color information
in images, and is usually an aspect of electronic noise is removed to eliminate obscures
in original information.
6.3 Feature extraction:
The feature is stated as a piece of information about content of image. Feature
Extraction aims to fine tune specific structures in the image such as points and edges.
6.4 Recognition:
Image recognition is a technique for capturing, processing examining and sympathizing
images. It is an ability to notice fine features in a digital image.
6.5 Text to Speech:
The good quality of text image is finally converted into audio with the help of text-to-
speech (TTS) engines, which can be heard through headphones.

7.Hardware Implementation

The proposed smart reader glass system includes the hardware components, which
include Raspberry Pi, battery package, rectifier unit, micro camera and microphone as
shown in Figure 3.
The Raspberry Pi is a small computer that plugs into the monitor and connects to a
 S. Anbarasi et al. / Smart Reader Glass for Blind and Visually Impaired 9
keyboard. Raspberry Pi 3 is an active compact sized electronics board which is more
powerful with 10x faster processing speed than first generation model. In addition it is
equipped with wireless Bluetooth facilities with HDMI or VGA port. Raspberry Pi 3 is
powered up with battery setup connected through micro USB cable. The image-
processing unit, the software program modules and python can be installed in the
Raspberry Pi.

Figure 3. Hardware module equipped with spectacle and headphone

The Micro camera is fitted on the spectacle as shown in Figure 4. The input power
supply from battery is connected through the USB port and a regulated output voltage
of +5V with output current of 100 mA fed to the Raspberry Pi and the audio output can
be accessed through the headphones.

Figure 4. Hardware module with Raspberry Pi

8.Summary and Conclusion

The proposed project - Smart reader glass for blind and visually impaired people
mainly aimed to assist the blind and visually impaired people to read the documents in
the form of text book, mobile screen or computer screen. Even though some of the
glasses are available in the market they are too costly and not affordable to common
people. Our proposed project is designed with simple components which can be placed
in shirt pocket. Hence this product is user friendly and will be available at very low
price and
92 S. Anbarasi et al. / Smart Reader Glass for Blind and Visually Impaired People

compact size. The audio output can be heard using normal microphones used for
mobile phone. The rigid assembling of hardware components ensure the lifetime of the
proposed smart glass. Really, this smart reading spectacles will be an optimal choice for
visually impaired peoples.

References
[1] Mohamed, S. A. E., Hassanin, A. S., & Othman, M. T. Ben. (2014). Educational System for the Holy
Quran and Its Sciences for Blind and Handicapped People Based on Google Speech API. Journal of
Software Engineering and Applications, 07(03), 150–161. https://doi.org/10.4236/jsea.2014.73017
[2] Sabab, S. A., & Ashmafee, M. H. (2016). Blind Reader: An intelligent assistant for blind. 2016 19th
International Conference on Computer and Information Technology (ICCIT), 229–234.
https://doi.org/10.1109/ICCITECHN.2016.7860200
[3] Sarkar, R., Das, S., & Roy, S. (2013). SPARSHA: A Low Cost Refreshable Braille for Deaf-Blind
People for Communication with Deaf-Blind and Non-disabled Persons (pp. 465–475).
https://doi.org/10.1007/ 978-3-642-36071-8_37
[4] Sharma, A., Devi, S., & Verma, J. K. (2020). Braille Book Reader using Raspberry Pi. 2020
International Conference on Computational Performance Evaluation (ComPE), 841–843.
https://doi.org/10.1109/ ComPE49325.2020.9200110
[5] K, S. G., & T, S. (2017). Real Time Implementation of Optical Character Recognition Based TTS
System using Raspberry pi. International Journal of Advanced Research in Computer Science and
Software Engineering, 7(7), 149. https://doi.org/10.23956/ijarcsse/V7I7/0117
[6] Ali Hassan, E., & Tang, T. B. (2016). Smart Glasses for the Visually Impaired People (pp. 579–582).
https://doi.org/10.1007/978-3-319-41267-2_82
[7] Alkhalaf, K. (2014). OCR-Based Electronic Documentation Management System. International Journal
of Innovation, Management and Technology, 5(6). https://doi.org/10.7763/IJIMT.2014.V5.560
[8] Bhui, N., Prasad, D., Sinha, A., & Kuila, P. (2021). Design of an Automatic Reader for the Visually
Impaired Using Raspberry Pi (pp. 175–188). https://doi.org/10.1007/978-981-15-7533-4_14