Smart Cap: A Sensor Based Low-Priced Assistant for the Blind

and Visually Impaired People

LICENSE

CC BY-NC-SA 4.0

SUBMISSION DATE / POSTED DATE

29-07-2022 / 05-08-2022

CITATION

Chakraborty, Arka (2022): Smart Cap: A Sensor Based Low-Priced Assistant for the Blind and Visually
Impaired People. TechRxiv. Preprint. https://doi.org/10.36227/techrxiv.20405838.v1

DOI

10.36227/techrxiv.20405838.v1
 Smart Cap: A Sensor Based Low-Priced Assistant
for the Blind and Visually Impaired People
Arka Chakraborty
Electrical Engineering
Indian Institute of Technology Kanpur, India
[email protected]

Abstract: The ability to perceive light is termed as mimic’s the sound produced and received by
vision. It can be referred to as one of the most animals and the generated information is
important sensory organs in most of the animals. We accordingly conveyed to the user after processing
use our visionary capability to understand and
the signal received with the help of a microcontroller
perceive our environment and take decisions based on
or microprocessor (we are using ATmega328P
these sensory inputs. Unfortunately, a large section of
our population faces some or other sort of visionary microcontroller present in Arduino Uno board).
defects. According to WHO around 2.2 billion people
II. LITERATURE SURVEY
worldwide [29] has some sort of visionary impairments
out of which iapb data suggests 43 million are Many ideas have been proposed to assist blind and
completely blind [30]. According to ncbi [31] India
visually impaired people to help them visualise their
alone has around 20 percent of world’s total visually
environment and help them in mobility. The work on
impaired population. In this paper we are discussing
an approach that will use ultrasonic sensors to detect assistive technology for the blind people has mainly
objects around and will inform the user about the three approaches: -
impending or stationary objects around him via
vibrational actuators. It is very cheap compared to
1. Mobility Assistance: This approach includes
other alternatives and is highly integrable with other scanning the user’s environment and conveying the
information back to the user either in the form of
proposed ideas.
audio messages or tactical feedbacks [1]-[17].
Keywords – WHO (World Health Organization), iapb,
2. Computer Access: These approach-based
ncbi, ultrasonic sensors, vibrational actuators.
solutions include voice generators and synthesizers,
I. INTRODUCTION braille output terminals and screen magnifiers [18],
[19].
The advancement of embedded systems and IoT has
changed almost every aspect of human day to day 3. Information Transmission: This approach mainly
life. They are widely used in automating houses, focuses on optical character recognition and
cars, watches, garden, and crop irrigation systems information rendering from either 2D or 3D scenes
and so on. Combined with cloud platforms they are [20]-[25].
revolutionizing every part of our life in every Some of the notable works and ideas of different
possible direction. In this paper we will be authors are listed below:
discussing a noble way to use proximity sensors
(ultrasonic sensor) to measure distance and use it to In [1]-[12] and [18] attempts have been made to
alert the blind users about nearby obstacles and integrate proximity sensors like ultrasonic and PIR
objects. Animals like bats and dolphins use a sensor along with other sensors such as water sensor
technique called echo location to get idea about their and infrared sensor in white cane or walking sticks.
surroundings. They send sound waves mainly Actuators like buzzers or sound modules have been
ultrasonic into their surroundings and listen for the used to convey the information obtained regarding
reflected sound wave called echo. From the received the environment to the blind user. Another approach
sound wave, they get idea about their surroundings, involves integrating the sensors in the gloves of the
prey’s, predators, or any other objects. This user [13]-[14]. This helps the user to understand the
technique is used by many animals whose natural nature of the object they are about to touch. Smart
ecosystem is either dark or is devoid of sufficient shoes are also proposed in [16] and [28] where
light. The idea behind this work is to use this proximity sensors are attached to the shoe of the user
technique to give the perception of objects by for detecting nearby obstacles. In [25] a smart cap is
detecting them with sounds (using ultrasonic sensor) proposed where the sensors are integrated in a hat
and informing the user about them with actuators and the information obtained is processed using a
like vibrational actuator. The ultrasonic sensor microprocessor.
Integrating sensors in a hat or a cap has an D. Limited performance: Many of the above models
advantage, as the sensors are almost present at the proposed by different authors uses sensors that
same height of the eyes, so the distance calculated either has limited range, like IR sensor or is unable
by the proximity sensors will be somewhat closer to to predict the direction of the detected obstacle like
what he would have seen. This allows for an easier in case of single ultrasonic sensor-based models.
correlation and helps the user to adapt to the
proposed device faster. Thus, we have developed Thus, to remove these limitations while maintaining
our proposed work on a hat. performance in terms of object detection range and
small response time while ensuring low cost, we
The different models and approaches that are have proposed and implemented the model
mentioned above tries to find the solution of discussed in section Proposed System.
providing the sense of distance to the user in
different ways. But implementation and deployment III. SYSTEM COMPONENTS DESCRIPTION
of these models have some limitations. Some of To design the proposed system, some prior
these issues are discussed below - important information about some of the
A. Cost: The cost of the sensors (like GPS) and components used are essential. They are discussed
microcontroller/microprocessor boards used in brief below.
(Raspberry Pi) increases the overall cost of the A. Ultrasonic Sensor – The ultrasonic sensor used
system. Now in South Asian and African countries here is HC-SR04 (Figure 1). The working of this
like India, Bangladesh, Sri Lanka, etc. cost becomes sensor is discussed briefly for better understanding.
an important factor while considering deployability
of a device. Most of the cases of visually impaired A HC-SR04 sensor consists of four pins, they are
people are present in poorer section of the society VCC (5V supply), Trigger, Echo and Ground (0V
and thus it becomes impossible for them to buy such supply). These modules can measure distance from
expensive equipment’s and bear the maintenance 2 cm to 400 cm with accuracy ranging up to 2mm.
cost (hardware maintenance, cloud services fee). In a HC-SR04 module we first make the Trigger pin
Thus, more often they either depend on government low for 5 microseconds and then send a 40 KHZ
or NGOs for such devices or spend their lives in trigger pulse signal from the sensor by keeping the
darkness. So, we aimed at designing a module that trigger pin high for 10 microseconds. Then we wait
will be cost effective and will be deployable at a for the echo of the sent signal and store the duration
large scale with low maintenance cost. of the echo to return in a variable say ‘t’. Now as the
time t includes both to-and-fro journey time, we find
B. Real-Time operations: Most of the systems that the net distance of the object (d) from the source by
uses machine learning algorithms for image and face the formula
recognition are deployed in microprocessor-based d = t*0.0343/2 cm
boards like Raspberry Pi (as the memory capability
and speed of operation of microprocessors are more Here 0.0343 is the speed of the sound in
compared to that of microcontrollers) as in [12] and cm/microsecond. By this method we can calculate
[25]. But this raises the issue of non-real time the distance of the obstacle from the sensor.
operations. As microprocessors are used, the time
required for execution and processing of the input
data is not fixed and is variable. There may be cases
when by the time the output is generated the object
has already hit the user or the object it recognised
has already passed. This becomes an important
drawback of using this technology for
implementation. Thus, real-time operations are very
important while constructing the device as quick
results (within fixed duration) are required for better
performance and safety of the user. Figure 1 – HC-SR04 ultrasonic sensor
C. Lack of proper internet speed and access: In B. Bluetooth Module – The Bluetooth Module used
countries like India the ground reality remains, lack is HC-05. The Bluetooth module consists of two
of proper internet speed. Thus, deployment of modes namely, Data Mode (For transmission of
devices that are highly dependent on cloud services data) and AT Mode (For changing default
for proper functioning are not desirable. configurations). The Bluetooth module can be
operated in either master configuration or slave echo signal. Once the echo is received it transmits
configuration. By default, most of the modules are the data to the Arduino board, which then calculates
in slave configuration. In our model we will require the distance of the obstacle from the HC-SR04
one master configuration module (sender) and one sensors. The distance obtained from the four
slave configuration module (receiver). To enter directions are then sent to another Arduino Uno
master configuration, we use the AT command board connected to the wrist band via paired
mode with suitable configuration adjustments. Bluetooth modules (HC-05). As Arduino Uno
supports only serial communication, we cannot send
the four data’s parallelly via four separate channels
(Arduino Mega supports parallel communication,
but Bluetooth uses serial communication, so we are
not using Arduino Mega). Thus, we need to find a
way to send the data serially while maintaining a
label as to which data is from which sensor (namely
front, rear, left or right) and should trigger which
vibration module.

Figure 2 -HC-05 Bluetooth module

IV. PROPOSED SYSTEM

Figure 5 – Smart cap (Outer view)

Figure 3 -Data Sending Module (Master Module)

Figure 6 – Smart cap (Internal view)

The proposed model consists of a Sender Module The receiver module is a wrist band (Figure 7)
(Figure 5 and 6) and a Receiver Module (Figure 7). consisting of an Arduino Uno module, a Bluetooth
The Sender Module comprises of an Arduino Uno module (HC-05), wires, power source and four
board, an HC-05 Bluetooth module, power source, mechanical vibrational modules for generating a
wires and 4 ultrasonic HC-SR04 sensors fitted in sense of perception. The four vibrational modules
four directions perpendicular to one another in a are placed perpendicular to each other and are
cap/hat. The four directions are front, rear, left and positioned in contact with the skin of the user for
right of the cap. All the ultrasonic sensors are proper sensing of the vibrational energy. To achieve
connected to an Arduino Uno board. The HC-SR04 serial communication, we took the four individual 8-
sensors sends the trigger signal and waits for the bit data from the four ultrasonic sensors and form a
single 32-bit data by bit shifting and send the 32-bit
data via paired Bluetooth modules. Similarly, in the V. COST ANALYSIS AND COMPARISON
receiver side we decode the four 8-bit data’s one
after another from the received 32-bit data. In this Components Approximate Cost (Rs.)
way the knowledge of origin of data is preserved that Arduino Uno 300
in turn eases the process of allocation of data to the Arduino Mega 1,700
Arduino Nano 500
vibration modules.
Raspberry Pi 4 6,200
Now before sending the data’s, the two Bluetooth Ultrasonic Sensor 50
modules are needed to be paired with one another. Bluetooth Module HC-05 200
To achieve this, we use AT commands. The Cap 140
Bluetooth associated with the sender Arduino is White Cane 600
configured to be the master Bluetooth module and IR Sensor 60
Bluetooth associated with the receiver Arduino Water detector 60
module is configured to be the slave Bluetooth Vibrational module 30
module. After successful pairing transmission of the Voice Record and 500
Playback Mini Board
data is attained.
ARP33A3 900
Now, in the receiver Arduino, after decoding the GPS module 320
data obtained, we process the data for necessary GSM module 700
action to be taken. We set a pre-defined threshold 4 channel remote 300
value for the distance measured and once the Webcam 2,000
received data crosses the threshold value, the Power bank 1,500
corresponding vibration module will turn ON. Battery 25
Earphones 600
microphone 2,710
Shoes 1000
Glasses 400
Gloves 400
PIC microcontroller 1000
MQ2 gas sensor 100
ESP8266 (Node MCU) 190
buzzer 30
4 channel remote 300
RF-TX module 100
Vibration motor 170
Push button 10
switch 30
Wi-Fi module 100
Motion sensor 60
Power relay module 70
Figure 7 – Receiver Module (Wristwatch) Flex sensor 300
Lidar sensor 3,700
Now to further enhance the working of the system LCD display 120
the time for which the vibration module will stay ON Servo Motor 200
after starting to vibrate can be made inversely Motor driver 120
proportional to the distance of the object in PIC16F87XA MCU 770
consideration. Thus, longer the vibrational pulse, LED 2
closer is the object from the user. In this way the user Bluetooth module HC-06 300
will be able to understand and perceive any obstacle DC Jack 20
both moving and stationary in any of the four Earphone 300
directions. He will also get a relative sense of the RS-232 50
distance of the obstacle from him by the duration of Speaker module 170
the vibrational pulse. With practice the user will be Flame sensor 70
able to perceive the distance more accurately. Using
of white cane along with this device is recommended Table 1 – Cost analysis of individual components
for better safety.
 Model and Reference Components Approximate Comment
Cost (in Rs.)
Smart Cane [1] Ultrasonic sensors (3), MQ2 gas sensors, 3,300 The system is bulky, costly, prone
Arduino Mega, esp8266, GPS module, buzzer, to internet speed fluctuations and
vibrator modules, cloud maintenance charges, needs cloud maintenance charges
cane, battery
Smart cane [3] Ultrasonic sensor, PIC, white cane, water 2,200 The system is bulky, costly, prone
detector, earphones, LED, buzzer, battery to internet speed fluctuations
Smart Cane [4] Arduino Uno, ultrasonic sensors (2), water 2,500 Costly and gives sense of only
sensor, GPS module, GSM module, sim, 4 one direction
channel remote, buzzer (2), cane, battery
Smart Cane [5] Ultrasonic module, Arduino Uno, Vibration 2,200 The system is costly and prone to
motor, buzzer, GPS module, GSM module, cane internet speed fluctuation with
limited operation range
Smart Cane [7] Ultrasonic, soil moisture, infrared sensors, RF, 2,600 Bulky with considerable
GPS-GSM module, buzzer, vibration motor, maintenance charges due to use of
push button, speaker, Arduino uno, switch, cane GPS and GSM modules and
limited operation range
Smart walking stick [8] Arduino Uno, soil moisture, ultrasonic, flame 2,000 The system is very heavy and
sensor, audio playback, GPS, Wi-Fi module, bulky
push button, DC vibrator motor, piezo buzzer,
mobile, cloud maintenance charges, cane
Smart stick [9] Arduino Uno, infrared sensor, water sensor, 1,300 The system is cheap but provides
ultrasonic sensor, buzzer, speaker, vibrator information about single direction
module, cane
Smart stick [10] Arduino Uno, switch, vibrator module, 2,100 Bulky and prone to internet
ultrasonic sensors (5), buzzer, GPS module, fluctuations
motion sensors, text to speech converter Wi-Fi
Module, cane
Navigation gadget based on Voice Record and Playback Mini Board, 2,100 Short range of operation,
IoT [11] ultrasonic sensor, Atmega328, GPS, white cane,
battery, earphone, Rs-232, speech to text
conversion
Smart Cane [12] Raspberry Pi, GPS module, ultrasonic, buzzer, 7,400 Costly, non-real time operation
vibration motor, cloud maintenance cost, cane and prone to internet fluctuations
Smart Glove [13] Arduino uno, ultrasonic, flame sensor (KY- 1,600 Inconvenient to wear such bulky
026), GSM, buzzer, gloves gloves all the time
Smart Glove [14] Arduino Mega, Uno, voice module (ARP 3,400 Costly, heavy, and bulky for
33A3), RF-TX module, power relay module, regular use
flex sensor
Smart Shoes [16] Arduino uno, ultrasonic sensor, shoe, glasses, 2,100 Costly and prone to internet speed
battery, vibrating servo motor, buzzer fluctuation
Smart Torch [17] LIDAR, LCD display, RF transmitter, 4-wheel 5,000 The system is highly costly
motors, buzzer, motor driver, PIC16F87XA
MCU
Smart Cane [18] Buzzer, ultrasonic, Arduino nano, LED, 6,000 The system is highly costly and
Bluetooth (HC-06), DC jack, mobile has limited range of operation
Smart Cap [25] Cap, earphone, microphone, speaker, raspberry 10,000 Highly expensive, heavy, non-real
pi4, power bank, webcam time operation with considerable
maintenance charges.
Smart cap [proposed model] Arduino Uno (2), cap, ultrasonic sensors (4), 1,200 Long range of detection, direction
vibrational modules (4), Bluetooth modules (2), of obstacle is known, relatively
batteries (2) low price, low maintenance cost,
comparatively light weight

Table 2 - Cost comparison of different models*

*The prices mentioned are in Indian Rupees and may vary slightly from time to time
It can be observed from Table 2 that techniques emergency messages to pre-selected people. The
proposed in [1], [12], [14], [17], [18] and [25] are module developed is also optimised for faster and
considerably expensive to be implemented in large real time performance and an average response time
scale. In particular, the techniques proposed in [4] of 0.8 seconds is achieved after optimisation. The
and [15] are claimed to be less expensive in proposed module is successfully implemented and
comparison to other alternatives. However, it can be tested by blind-folded candidates.
seen that our proposed technique is much cheaper
compared to the above-mentioned techniques. VIII. REFERENCES

[1], [8], [10] and [12] are dependent on internet [1] Z. Saquib, V. Murari and S. N. Bhargav,
availability for proper performance. However, in "BlinDar: An invisible eye for the blind people
countries where proper internet infrastructure in making life easy for the blind with Internet of Things
remote areas still remains an issue, the above (IoT)," 2017 2nd IEEE International Conference on
techniques would face serious limitations in Recent Trends in Electronics, Information &
providing optimal performance. [4], [5], [7] and [11] Communication Technology (RTEICT), 2017, pp.
also uses GPS modules and thus require tri satellite 71-75, doi: 10.1109/RTEICT.2017.8256560.
interlocking for working. Thus, they are not [2] Yiting Yi and Lunfu Dong, "A design of blind-
effective in indoor operation or in bad weather guide crutch based on multi-sensors," 2015 12th
where receiving proper satellite signal is an issue. International Conference on Fuzzy Systems and
So, it is better to avoid these components when Knowledge Discovery (FSKD), 2015, pp. 2288-
reliable and speedy operation is desirable in the 2292, doi: 10.1109/FSKD.2015.7382309.
developed system. [3] Wahab, Mohd Helmy Abd, et al. “Smart cane:
Assistive cane for visually-impaired people.”
[9] and [13] are cheaper but limits in terms of range International Journal of Computer Science Issues,
of operation. [9] only provides the information about July 2011
presence of nearby objects but does not provide the [4] M. Ashrafuzzaman, S. Saha, N. Uddin, P. K.
information about the direction of the obstacle due Saha, S. Hossen and K. Nur, "Design and
to use of single proximity sensor. Our technique on Development of a Low-cost Smart Stick for Visually
the contrary provides a better sense of direction and Impaired People," 2021 International Conference on
distance from the object due to use of multiple Science & Contemporary Technologies (ICSCT),
proximity sensors. 2021, pp. 1-6, doi:
10.1109/ICSCT53883.2021.9642500.
Some of the above developed techniques are also
[5] S. Karkuzhali, A. Mishra and M. S. Ajay,
heavy and bulky, thus will also be impractical to
"Design of Neurocane for Blind Community Using
deploy without proper optimisation.
IoT Device," 2020 International Conference on
In comparison to other techniques our model is more Computer Communication and Informatics
cost effective, compact, lighter, immune to signal (ICCCI), 2020, pp. 1-7, doi:
fluctuation and also provides better sense of 10.1109/ICCCI48352.2020.9104075.
direction and distance of the object detected. [6] E. Cardillo, C. Li and A. Caddemi, "Empowering
Blind People Mobility: A Millimeter-Wave Radar
VII. CONCLUSION Cane," 2020 IEEE International Workshop on
Metrology for Industry 4.0 & IoT, 2020, pp. 213-
The main objective of this work is to propose and
217, doi:
design a device that will help the blind and visually
10.1109/MetroInd4.0IoT48571.2020.9138239.
challenged people to perceive there surrounding by
[7] V. Kunta, C. Tuniki and U. Sairam, "Multi-
using ultrasonic sensors. The idea behind the
Functional Blind Stick for Visually Impaired
developed device is to mimic the echo location
People," 2020 5th International Conference on
technique used by many nocturnal and aquatic
Communication and Electronics Systems (ICCES),
animals. We also tried to keep the cost of the system
2020, pp. 895-899, doi:
low with almost no maintenance charges. We made
10.1109/ICCES48766.2020.9137870.
it independent of cloud services to address the poor
[8] R. S. Krishnan, K. L. Narayanan, S. M. Murali,
network issue faced in many rural parts. However, if
A. Sangeetha, C. R. Sankar Ram and Y. H.
required GPS module can also be integrated in the
Robinson, "IoT based Blind People Monitoring
system with little increase in the cost that will enable
System for Visually Impaired Care Homes," 2021
real-time monitoring of the user’s location.
5th International Conference on Trends in
Emergency panic switches can also be interfaced in
the wrist module that will generate and send
Electronics and Informatics (ICOEI), 2021, pp. 505- InternationaFl Conference on System, Computation,
509, doi: 10.1109/ICOEI51242.2021.9452924. Automation and Networking (ICSCAN), 2020, pp.
[9] V. Natarajan, Y. M and A. Canessane, "IoT 1-6, doi: 10.1109/ICSCAN49426.2020.9262365.
based Smart Stick with Automated Obstacle [18] M. A. Khan et al., "An Android-based Portable
Detector for Blind People," 2022 6th International Smart Cane for Visually Impaired People," 2021
Conference on Trends in Electronics and IEEE 15th International Conference on Application
Informatics (ICOEI), 2022, pp. 536-540, doi: of Information and Communication Technologies
10.1109/ICOEI53556.2022.9777174. (AICT), 2021, pp. 1-6, doi:
[10] SathyaNarayanan E, Gokul Deepan D, Nithin B 10.1109/AICT52784.2021.9620268.
P and Vidhyasagar P, "IoT based smart walking cane [19] G. M. Siddesh, S. Manjunath and K. G.
for typhlotic with voice assistance," 2016 Online Srinivasa, "Application for assisting mobility for the
International Conference on Green Engineering and visually impaired using IoT infrastructure," 2016
Technologies (IC-GET), 2016, pp. 1-6, doi: International Conference on Computing,
10.1109/GET.2016.7916687. Communication and Automation (ICCCA), 2016,
[11] N. S. Mala, S. S. Thushara and S. Subbiah, pp. 1244-1249, doi: 10.1109/CCAA.2016.7813907.
"Navigation gadget for visually impaired based on [20] V. Kushnir, B. Koman and V. Yuzevych, "IoT
IoT," 2017 2nd International Conference on Image Recognition System Implementation for
Computing and Communications Technologies Blind Peoples Using esp32, Mobile Phone and
(ICCCT), 2017, pp. 334-338, doi: Convolutional Neural Network," 2019 XIth
10.1109/ICCCT2.2017.7972298. International Scientific and Practical Conference on
[12] D. S. AbdElminaam, I. A. -E. Ahmed and F. Electronics and Information Technologies (ELIT),
Sakr, "SCBIoT: Smart Cane for Blinds using IoT," 2019, pp. 183-187, doi:
2022 2nd International Mobile, Intelligent, and 10.1109/ELIT.2019.8892289.
Ubiquitous Computing Conference (MIUCC), 2022, [21] J. -E. Kim, M. Bessho, N. Koshizuka and K.
pp. 371-377, doi: Sakamura, "Mobile applications for assisting
10.1109/MIUCC55081.2022.9781728. mobility for the visually impaired using IoT
[13] J. W. Simatupang and B. R. Prasetya, infrastructure," Proceedings of 2014 TRON
"Embedded Smart Glove using Ultrasonic and Symposium (TRONSHOW), 2014, pp. 1-6.
Flame Sensors for Helping Visually Impaired [22] H. Hesham, L. Khalil, A. Hafez and M.
People," 2021 International Conference on Radar, Hussein, "Design, Simulation, and Implementation
Antenna, Microwave, Electronics, and of connected IoT for Disabled People," 2022 2nd
Telecommunications (ICRAMET), 2021, pp. 115- International Mobile, Intelligent, and Ubiquitous
119, doi: 10.1109/ICRAMET53537.2021.9650494. Computing Conference (MIUCC), 2022, pp. 279-
[14] K. P. k. R. Shankar, E. K. Sawant and Y. 283, doi: 10.1109/MIUCC55081.2022.9781761.
Punarvit, "Arduino Based Smart Gloves for Blind, [23] S. Choudhary, V. Bhatia and K. R. Ramkumar,
Deaf and Dumb Using Sign and Home Automation "IoT Based Navigation System for Visually
Mode," 2021 International Conference on Advances Impaired People," 2020 8th International
in Technology, Management & Education Conference on Reliability, Infocom Technologies
(ICATME), 2021, pp. 244-248, doi: and Optimization (Trends and Future Directions)
10.1109/ICATME50232.2021.9732739. (ICRITO), 2020, pp. 521-525, doi:
[15] S. Durgadevi, C. Komathi, K. 10.1109/ICRITO48877.2020.9197857.
Thirupurusundari, S. S. Haresh and A. K. R. [24] R. Fernandes, A. P. Rodrigues and K. B.
Harishanker, "IOT Based Assistive System for Sudeepa, "IoT based Smart Security for the Blind,"
Visually Impaired and Aged People," 2022 2nd 2018 Fourteenth International Conference on
International Conference on Power Electronics & Information Processing (ICINPRO), 2018, pp. 1-4,
IoT Applications in Renewable Energy and its doi: 10.1109/ICINPRO43533.2018.9096750.
Control (PARC), 2022, pp. 1-4, doi: [25] Amey Hengle, Atharva Kulkarni, Nachiket
10.1109/PARC52418.2022.9726626. bavadekar, Niraj Kulkarni, Rutaja Udyawar. “Smart
[16] T. Chava, A. T. Srinivas, A. L. Sai and V. Cap: A Deep Learning and IoT Based Assistant for
Rachapudi, "IoT based Smart Shoe for the Blind," the Visually Impaired.” IEEE,2020.
2021 6th International Conference on Inventive [26] M. Zhou, W. Li and B. Zhou, "An IoT System
Computation Technologies (ICICT), 2021, pp. 220- Design for Blind," 2017 14th Web Information
223, doi: 10.1109/ICICT50816.2021.9358759. Systems and Applications Conference (WISA),
[17] S. Kalpana, S. Rajagopalan, R. Ranjith and R. 2017, pp. 90-92, doi: 10.1109/WISA.2017.72.
Gomathi, "Voice Recognition Based Multi Robot [27] Amjed S. Al-Fahoum, Heba B. Al-Hmoud. And
for Blind People Using Lidar Sensor," 2020 Ausaila A. Al-Fraihat. “A Smart Infrared
Microcontroller-Based Blind Guidance System,” Cloud Computing, Data Science & Engineering
Active and Passive Electronic Components, vol. (Confluence), 2021, pp. 552-556, doi:
2013, p.7, 2013 10.1109/Confluence51648.2021.9377120.
[28] S. Rao and V. M. Singh, "Computer Vision and [29] WWW.WHO.int
Iot Based Smart System for Visually Impaired [30] www.iapb.org
People," 2021 11th International Conference on [31] WWW.pubmed.ncbi.nlm.nih.gov