Arduino Based Customized Smart Glasses For The Blind People
Arduino Based Customized Smart Glasses For The Blind People
R.M.K College of Engineering and R.M.K College of Engineering and R.M.K College of Engineering and
Technology, Anna University Technology, Anna University Technology, Anna University
Chennai, India Chennai, India Chennai, India
[email protected] [email protected] [email protected]
Abstract— For all human beings’ vision is primary information and sends corresponding signal to buzzer.
requirement to survive in fast-changing environment. The corresponding buzzer sound will help the subjects to
S cientific glasses are there to improve the eye sight of identify the objects or obstacles accordingly.
visually impaired people. But for blind people there is a
need for smart glass to guide them to be independent. In This paper is organized as follows: Section II,
this paper, we have proposed a method to support visually presents the related works involved in guiding the
challenged people. The existing methods are available to visually challenged people. The proposed customized
identify obstacles, stair cases, water proximity separate ly as solution is presented in Section III. In Section IV,
well as combinedly. We have extended the existing smart aid experimental results are discussed. Finally, conclusions
techniques for obstacle detection with the fire detection and are given in Section V.
also the background detection. We have developed a low-
cost solution using the input and output sensors connected II. LITERATURE SURVEY
through Arduino board. The environment in front of the There exist a number of techniques to detect the
subject is detected by sensors and then the signal will be obstacles using Arduino board. Some of them
sent to the processor. The processor will respond to the incorporated the different types of techniques to develop
input and sends the corresponding signal to the output
transducer. The buzzer sound or beep sound from the
the smart glasses as a single solution for detecting
different types of obstacles.
output transducer will alert the user accordingly.
Aman et al. [1] reviewed the advances in smart glass
Keywords— Smart aid, Visually challenged, Sensors, technologies to guide the blind people in navigation
Arduino tasks. Many of them used the deep learning techniques
I. INTRODUCTION with stereo cameras to calculate the distance between the
blind people and the obstacles. Amreen Khanam et al. [2]
In today’s world, the number of visually impaired reviewed the different types of techniques like RFID,
people is increasing day by day over the past decades. As Virtual senses, smart cane etc to give support to visually
reported by World Health Organization (WHO), about impaired people. They also extended the existing
285million people worldwide are estimated to be visually technique by integrating a camera in it. Andres A.
impaired. And also, the 90% of them still live-in low
level of income. “The idea behind this work is to come Diaz-Toro et al. [3] used stereo cameras for depth
up with smart and low-cost solution to help the visually measuring and the inertial sensors for segmenting the
impaired to lead independent lives”. Smart glasses for floor. They have developed a navigation algorithm for
visually impaired people are wearable devices that have mapping the instruction about free space into a motion
sensor input components, a processing unit and an output algorithm, which guides the blind people in an unknown
transducer. The motivation behind this work is to design environment without collisions. Ankitha Bhuniya et al.
and develop a customized solution for visually impaired [4] introduced the concept of Interactive Voice Response
people. And also, we have incorporated the multiple System (IVRS) to identify the direction of obstacle. It
sensors into single system to enhance the existing employs four ultrasound transducers each at front, back,
devices. left and right along with a micro controller. Arvitha et al.
[5] developed the technology to detect obstacles, stair
The proposed model of smart glass for visually cases, and water puddles using ultrasound and IR
impaired enhance the features of the existing obstacle proximity sensors and controlled by Arduino UNO.
detection models. It is an embedded system integrating
the pair of ultrasound transducer, flammable gas or Bineeth Kuriakose et al. [6] discussed the limitations
smoke detector and Light Dependent Resistor (LDR) to of the existing systems used in navigation of blind
identify the objects like obstacles, fire, smoke and people. They listed the three main reasons: they are
darkness. The information obtained from sensors is given portability and comfortability of the device, adaptation
to the Arduino processor. Arduino UNO process the time to device and user learning and adaptable feedback
options. They have classified the tools and technologies crowd at top atmosphere so placing such sensor in height
available for blind people into five categories: 1. Visual increase the performance.
imaginary systems, 2. Map based systems, 3. Non visual
systems, 4. Smart phone-based applications, and 5. TABLE I. COMP ARISON T ABLE WITH EXISTING WORK
Systems with 3D sound. Freddy Poly et al. [7] have S. Author Method/ Features/ Hardware /
developed a VI spectacle and a mobile app. This work No. Name Technology O bjective Sensors
extended the existing object detection techniques with an used
ESP32 camera module and they have transferred images 1 Prototype Arduino obstacle, fire, ultrasound
by the camera using a Wi Fi connection to the mobile. of this UNO background transducer,
work detection smoke
These images were processed by YOLOv3 object detector,
detection module, and then the corresponding audio LDR
output was sent to the user through earphones. 2 Aman et deep calculate the stereo
al. learning distance cameras
Jinqiang Bai et al. [8] have developed the smart 3 Amreen RFID, calculate the camera
guiding glass for indoor environment using multi sensor Khanam Virtual distance
technology. It includes a depth sensor along with an ultra et al senses,
sound sensor to detect the small as well as the transparent smart cane
objects. Md. Khaled Hassan et al. [9] reviewed and 4 Diaz-Toro A depth stereo
et al navigation measuring, cameras,
presented the systems and the equipment’s to support the algorithm segmenting inertial
daily lives of visually impaired people. They have also the floor. sensors
surveyed the Convolutional Neural Networks designs of 5 Ankitha IVRS direction of 4 ultrasound
smart glasses technology. They have compared the Bhuniya obstacle transducers,
objectives of different works involved in smart glass for et al a micro
visually impaired people. They have also presented the controller
6 Arvitha et Arduino detect Ultrasound,
different types of sensors used in the detection of objects. al UNO obstacles, IR
Lavanya Dhanesh et al. [10] developed a portable stair cases, proximity
and water sensors,
model for visually impaired people. It measures the puddles
distance of the objects and sends the alerts in the form of 7 Freddy VI spectacle object ESP32
vibrations. It also includes the toggle switch, which alerts Poly et al and a detection camera
surrounding people if there is a necessity. Maghfirah Ali mobile app module,
et al. [11] presented a low-cost single board Raspberry Pi YOLOv3,
earphones.
design of smart glasses for visually impaired students to
8 Jinqiang multi sensor detect the depth
extract the text and to convert it in to an audio. They have Bai et al technology small, sensor, an
used SIMULINK software to process the image, the transparent ultra sound
Optical Character Recognition (OCR) software to extract objects sensor
the text from the image and the speech synthesizer to 9 Lavanya vibrations distance of toggle
produce the audio output from the processed text. Dhanesh the objects switch
et al
Rohit Agarwal et al, [12] presented an Arduino model 10 Maghfirah SIMULINK, extract the OCR
to detect obstacles by using a SONAR sensor. It will Ali et al Raspberry text and to
detect up to the range of 300m. Senthil Kumar et al. [13] Pi convert it in
to an audio
employed the CNN trans fer learning algorithms for the 11 Rohit Arduino detect SONAR
detection and also the classification of obstacles using the Agarwal obstacles sensor
MATLAB software. They have also presented a walking et al
shoe for detecting ground level obstacles. The shoe is 12 Senthil MAT LAB, Classification ultra sound
incorporated with an ultra sound sensor for obstacle Kumar et CNN of obstacles sensor,
detection, a mems sensor for detection of pit falls, and a al transfer mems
learning sensor
moisture sensor for detecting water stagnant, slippery algorithms
water. Sulaiman Khan et al. [14] depicted the 13 Sundar Raspberry recognise the Pi camera
contribution of north America in IoT based medical Srinivas Pi face
appliances. They also identified that an IoT based et al
navigation application is one of the best techniques to
guide the visually impaired people in indoor and outdoor
In this paper, we have presented an innovative
environments. Sundar Srinivas et al. [15] used the
integration of different techniques to detect the three
Raspberry Pi technology with a Pi camera to recognise
different types of surrounding environment. The key
the face. It detects objects of the range of 1000cm. contributions of the proposed idea are
The prototype of this work is compared with the
1) To design and develop a cost effective and reliable
existing systems in terms of methodology, features and
sensors used in the design and presented in table I. In product to the visually impaired people using
proposed customized model, three sensors are used in Arduino.
smart glass for different sensing purpose but in existing 2) The proposed smart glasses can improve the
smart glass obstacle sensor are mostly used which traveling experience of visually impaired people.
enhanced the performance of proposed design. In the
proposed model, smoke sensor is used in smart glass
instead of stick because at time of fire accident the smoke
BATTERY
LM393
(LDR)
5) MQ2 Flammable Gas and Smoke Sensor The details of components and their specifications are
Module: At last, the MQ2 sensor is used to detect the fire given in table II.
and the smoke. It will send signals to the Arduino nano, if
TABLE II. COMP ONENTS AND THEIR SP ECIFICATIONS
it detects the fire or smoke and is as shown in Fig. 7.
S. No Hardware Specification
3 LDR LM393
D. Pin description
Ultrasonic sensor trigger, echo pins connect to 2, 3
pins of Arduino. MQ2 smoke sensor with 4th pin of
Arduino. LDR with 5th pin of Arduino and 6th pin with
peizo electric buzzer are shown in Fig. 11.
Fig. 8. Light Dependent Resistor
IV. CONCLUSION
We have designed and developed a prototype model
for obstacle detection with extended features. With this
customization, it is very much useful to the visually
impaired people and those who often have to rely on
others. The computation is fast enough for the detection
and the display of obstacles using Arduino. Experimental
results show that the proposed smart glasses can improve
the traveling experience of visually impaired people. The
sensors used in this system are simple and with low cost,
Fig. 12. Prototype of proposed system making it possible to be widely used in the consumer
market. Our future implementation is to convert the
It accurately detected the smoke by using MQ2 output as a speech signal by replacing Arduino with
flammable gas and smoke sensor module and its working raspberry pi and using RFID technique.
module is shown in the Fig. 13. The obstacles are
detected by using a pair of ultrasound transmitters and its REFERENCES
working module shown in the Fig. 14. It detects the [1] Aman, Bhavesh B and Rahul C, “A Comprehensive Review of
obstacles within the range of 2 feet from user. It finds the Smart Glasses T echnology- Future of Eyewear”, T urkish Journal
darkness by using the LDR sensor and its working of Computer and Mathematics Education, vol. 12, no. 2, April,
module is shown in the Fig. 15. 2021, pp. 15-26.
[2] Amreen Khanam, S.K. Abdul Basith, Anjana Dinesh, Mohammed
Imaduddin and Noor Fathima, “A Survey on Smart Glasses: An
audio assistant for visually impaired”, Journal of Applied Science
and Computations, vol. 5, no. 12, pp. 96-101, December 2018.
ISSN NO: 1076-5131
[3] Andrés A. Díaz-T oro, Sixto E. Campaña-Bastidas, Eduardo F.
Caicedo-Bravo, "Vision-Based System for Assisting Blind People
to Wander Unknown Environments in a Safe Way", Journal of
Sensors, vol. 2021, Article ID 6685686, 2021.
[4] Ankita Bhuniya, Sumanta Laha, Deb Kumar Maity, Abhishek
Sarkar and Suvanjan Bhattacharyya, “Smart Glass for Blind
People”, AMSE JOURNALS-AMSE IIET A publication-2017-
Series: Modelling D; Vol. 38; No. 1; pp. 102-110, Nov.15, 2017
[5] Arvitha, M., Biradar, A.G., & Chandana, M., “ T hird Eye for
Visually Challenged Using Echolocation T echnology ”,
International Conference on Emerging Smart Computing and
Informatics (ESCI), pp. 391-395,
Fig. 13. Smoke detection 202110.1109/ESCI50559.2021.9397015.
[6] Bineeth Kuriakose, Raju Shrestha & Frode Eika Sandnes, “ T ools
and T echnologies for Blind and Visually Impaired Navigation
Support: A Review”, IET E T echnical Review, 2020, DOI:
10.1080/02564602.2020.1819893
[7] Freddy Poly, Dipak T iwari, Varghese Jacob, “VI Spectacle – A
Visual Aid for the Visually Impaired”, International Research
Journal of Engineering and T echnology, e-ISSN: 2395-0056, Vol.
7, no. 2, pp. 1153-1155, Feb 2020, ISO 9001:2008.
[8] J. Bai, S. Lian, Z. Liu, K. Wang and D. Liu, "Smart guiding
glasses for visually impaired people in indoor environment," in
IEEE T ransactions on Consumer Electronics, vol. 63, no. 3, pp.
258-266, August 2017, doi: 10.1109/T CE.2017.014980.
[9] Md. Khaled Hassan, Md. Rakibul Hassan, M.M. T anvir Ahmed,
Md. Sayeed Ahmed Sabbir, Md. Sabbir Ahmed, and Milon
Biswas, “A Survey on an Intelligent System for Persons with
Visual Disabilities”, Australian Journal of Engineering and
Innovative Technology, vol. 3, no. 6, pp. 97-118, 2021, DOI:
10.34104/ajeit.021.0970118.
Fig. 14. Obstacle detection [10] Lavanya Dhanesh, B. Surya Prakash, P. Sai Dinesh, S. Kavi
Kumar, K. Nivin, “ Smart Glass for Visually Impaired Persons”,