Proceedings of the Sixth International Conference on Inventive Computation Technologies [ICICT 2021]

IEEE Xplore Part Number: CFP21F70-ART; ISBN: 978-1-7281-8501-9

Automation of Noise Detection Using Internet

of Things

Dr.C.N.Van itha, K.L.Sridhar, R.Dhivakar,

2021 6th International Conference on Inventive Computation Technologies (ICICT) | 978-1-7281-8501-9/21/$31.00 ©2021 IEEE | DOI: 10.1109/ICICT50816.2021.9358628

Professor, Scholar, Scholar,

Department of CSE, Department of CTUG, Department of CTUG,
Kongu Engineering College, Kongu Engineering College, Kongu Engineering College,
Perundurai , Perundurai, Perundurai.
vanithacn.ctug@kongu.edu sridharkl.18bcr@kongu.edu dhivakarr.18bcr@kongu.edu

Abstract – With a significant increase in population in that, the person who speak loudly can be alerted with
modern cities, the noise pollution also increases at an an alert message. This alert message can be send to
unprecedented rate. This system presents a low cost the person personally. And the alert message can also
flexible and reliable automation system with be send to the person who is taking care of the
personal alert messages using Arduino along with
library. These things can be done by using the
internet connectivity and remote control over devices
by using a smart phone application. The proposed
Internet of Things. It is not possible for a wider area.
system uses Internet of Things[IoT] model to detect This is possible for only a limited area. It should be
human voices present in an enclosed area. The user very useful in library to maintain a silence for a long
can also receive the alert messages with the help of period of time. Another good thing about this system
a smart phone. This research work presents the is noise pollution can also be reduced.
automation of noise detection system using Arduino
Nano, battery, LCD panel and buttons. C In some working area, the coworker can
programming language has been used to program the make the loud noises and disturb you. To avoid those
functions of this application. This application can be things this method has been implemented. In recent
implemented by using a tool called Pick2kit and years, there are many researches and system about
PROTEUS . The proposed noise detector system can this automation of noise detection. Those researches
be used in library, office and classroom
and system results with the detection of noises only.
environments to identify the noisy people so that a
Fig.1 shows the recording device to record the
necessary action can be taken against them.
conversion. This should be very useful for the person
Keywords--- Noise detection, Arduino, Internet of Things, those who are using this.
automation.

I. INTRODUCTION

Noise pollution is an increasing issue in the

emerging modern world, where rapid population
growth may further increase the noise pollution.
Whenever people hear about Silence, the first thing
comes to the mind is a library. Since the library is the
place, where people should maintain utmost silence.
But not all the people, who come to the library
maintains the silence. To avoid these type of
activities and warn the people, who are making the
loud voice, a noise detector has been proposed in this
Fig.1 The recording device records the conversation
research work. This noise detector can detect the
of two peoples
voice of the person, who are making the voice louder
than allotted level and they can be continuously
monitored to take necessary action. In addition to

978-1-7281-8501-9/21/$31.00 ©2021 IEEE 184

Authorized licensed use limited to: Dedan Kimathi University of Technology. Downloaded on May 18,2021 at 15:08:25 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the Sixth International Conference on Inventive Computation Technologies [ICICT 2021]
IEEE Xplore Part Number: CFP21F70-ART; ISBN: 978-1-7281-8501-9

II. RELATED WORK mitigation approach required by the orthogonal

frequency division multiplying (OFDM). In the stage
The work that contains speech recognition first, to detect instances of impulsivity, deep neutral
system having two procedures called proposed noise network is used [5]. In the suppression stage, the
reduction. To identify the content of speech that detected IN is to alleviate the harmful effects. There
contains ambient noise, these systems can be used are two methods to demonstrate the DNN based
Noise reduction and threshold based noise detection. approach, 1.) IN models for testing and training, 2.)
This are the two stages in this system. Stage one Burstimpulsive environment with interleaving
determines the quantity of speech, second stage techniques.
consists of noise reduction [1]. The Quality of speech
recognition improved around 15% to 25%. In this Automation of city is the aim of ever smart
noisy environment, this system reduce the effect of city and it is provided with technological access. In
noise and improves quality of speech recognition. this paper, a Machine learning based application is
used for the detection of urban noise for the security
To generate a frame signal, frame signal of public and neighborhood safety. Various
generator is used and getting its time length from Algorithm of Machine learning by the urban noise
input signal. And to generate a reference signal from detection includes convolutional neural networks,
where the signal has been located, reference signal Long short term memory, decision tree random forest
generator has used to find a correct value between the naive Bayes with 311 data set [6]. For the ML
frame signal and reference signal correlation algorithms also this experiment applicable on mobile
calculator is used[2].to determine the frame signal devices it can be conducted and connected to system
includes periodic noises and to calculate the period of by sending requests.
periodic noises. To determine the phase shift amount,
range generator is used. Up to this, new sensing method for aero
engine fan noise test only proposed. In this paper,
Open plan offices have less costs for details of validities aero engine fan noise test method
construction, allows only important savings of space, and wind tunnel test. This experiment consists
to make communication between working people, uniform to simulate flow by the duet, to generate
improves working practice as same as exchange of different characteristics of noise, and an attention is
thoughts for this, those types of offices making focused on examination with reconstructed
wider. It includes the controlling of anthropic and association with accuracy and probability of success
electro mechanical noise. In that noise measurements [7]. The potential capability of this new fan mode
given by ventilation, heating and Air conditioning method of detection that are practically tested
[3]. To find the importance of variables, to identify facility.
the important characteristics. Number of Predictor
variables can be reduced by analysis the In the tiller-Huang transform, this letter
characteristics. Adoption of tool for many gives a speech enhancement method to mitigate the
applications are determined by high accuracy model. effects of impulsive noises selection and estimation
of noise is based on impulsiveness [8]. It conducts
In between the multiple microphones an speech experiments consists of five acoustic noises
audio record device has been implemented. It has with different index and non-stationarity degrees
been implemented to generate an output audio signal comes under signal to noise radius. Three algorithms
depends on conditions. If the presence of wind noise are defined in the evaluation contains spectral and
and other noise has been detected, the audio time domains. The best results from the proposed
recording device selects for each different frequency solution in terms of quality measures to the
and the lowest noises contained by an audio signal competitive methods.
and the combination of frequency to generate output
signal [4]. When there is no detection, the audio In Intensive care unit, Electrocardiogram
recording device needs the microphones whether wet signals that recorded are often corrupted by motion
or dry and selects audio signals based on conditions. and noise air facts, which treads false alarms includes
inaccurate detection. This system is an automated
In many communication applications, method to detect from ECG recordings. Irregular RR
removed of impulsive noise from received signal is intervals from the QRS complexes are identified, to
very important. This paper has two stages of

978-1-7281-8501-9/21/$31.00 ©2021 IEEE 185

Authorized licensed use limited to: Dedan Kimathi University of Technology. Downloaded on May 18,2021 at 15:08:25 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the Sixth International Conference on Inventive Computation Technologies [ICICT 2021]
IEEE Xplore Part Number: CFP21F70-ART; ISBN: 978-1-7281-8501-9

identify the QRS complexes have to focus on recognize voice commands, convert them to
detection algorithm. An algorithm is designed desired data coordination and data
between ECG and non ECG wave forms [9]. This transmission via and micro controller
method is to find accurately between segments (Arduino Uno) Receiving this signal by PC
contained real ECG forms. It was founded that Monitoring (IOT )then get a full control that
important reduction in false positive detection. works by voice commands. The software
Inaccurate detection of AF results without using it. consists of a Micro C language
programmable microcontroller. This system
To detect the noise of the wind, a processed is of low cost and flexible with growing
digitalized microphone signed has been used. In at variety of devices that can be controlled.
least one of the microphones, both the first and  Operating voltage is 5V. The input voltage
second are obtained. Reflection of conman input is
ranges from 7V to 12V.
done by the both signals and also involves
 USB Serial Converters allow new laptops to
temporarily or spatially. The process of first signal to
communicate safely with … Wide operating
define the first distribution samples [10,11].The
temperature range for harsh environments.
process of second signal to define the second
Turn Serial Devices into IoT-Connected
distribution has been calculated. An output indicates
Devices.
when there is wind noise [12].

III. PROPOSED WORK T ABLE II. NOISE LEVELS OF EVERYDAY SOUNDS

A voice model established by the system is Decibel Environment

according to voice characteristics of human,
10 decibel Sound during breathing
analyzing the input voice signal and proceed the
required features on this basis, it establishes the 20 decibel Whispering away from five feet
required template of the speech recognition. The
30 decibel Whispering nearby
computer is used in the identify process by the model
of the speech recognition to compare the voice 40 decibel Sounds in quiet library
template that has stored inside computer and the 50 decibel Sound while refrigerator is on
characteristics of the input voice signal. Search and 60 decibel Electric toothbrush
matching strategies to identify the optimal range of
the voice matches the template. 70 decibel Sound while using Washing
machine
T ABLE I. HARDWARE USED 80 decibel Alarm clock

Arduino Uno It is the controller board, 90 decibel Sounds in Subway train

which controls the other 100 decibel While using Machinery in
hardware. factory
110 decibel Sound while using Car horn
Step-Down Reduces 230V and give it
transformer to the other hardware. 120 decibel Siren
Led Display To display the title of the
project.
Voice Recognition Used to play the recorded The Table II represents the noise levels of everyday
Board voice. sounds that people are hearing in a modern day.
Power supply board It gives the power supply
A. METHODOLOGY
to the hardware.
Internet of Things (IOT) is currently an
 Table I displays the hardware used. emerging technology spread globally. IOT has
According to the definition of this template stepped in various fields comprising of Industries,
through the lookup table can be given the Government, academia, and still various research are
recognition results of the computer. Voice carried in this domain. Nowadays IOT plays a vital
commands based on the voice recognition role in the business sector, many glooming and
system (voice recognition unit ) system can blooming business depends on IOT & Industrial IOT.

978-1-7281-8501-9/21/$31.00 ©2021 IEEE 186

Authorized licensed use limited to: Dedan Kimathi University of Technology. Downloaded on May 18,2021 at 15:08:25 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the Sixth International Conference on Inventive Computation Technologies [ICICT 2021]
IEEE Xplore Part Number: CFP21F70-ART; ISBN: 978-1-7281-8501-9

Input:
Play the recorded voice in the voice
recognition board
Output:
START
VOICE DETECTION ALERT
1. Begin the process
2. Reduces the voltage from the 230V to
12V by Step down transformer
TRANSFORMER
3. Train the power supply board (230V)
4. p=a(5V) v(5V) l(5V) (p=POWER
SUPPLY BOARD)
5. t>230V(t=TRANSFORM ER)
POWER SUPPLY
6. a=5V(a=ARDUINO UNO) BOARD
7. v=5V(v=VOICE RECOGNITION
BOARD)
8. l=5V(l=LCD DISPLAY)
9. Connect the pins to the arduino ARDUINO
10. Code the arduino
11. #For the powersupply
12. Power supply board to vin
13. Power supply board to ground
14. Enable 5V to voice recognition board
VOICE No
15. LCD enable pin to 11 RECOGIN STATUS
16. LCD D4 pin to 2 ITION ZERO
BOARD
17. LCD D5 pin to 3
18. LCD D6 pin to 4
19. LCD D7 pin to 5
20. Liquid Crystal lcd(rs,en,d4,d5,d6,d7) Yes
21. int led=1
22. Code the arduino to voice recognition ALERT
board MESSAGE

23. Voice recognition to pin 13

24. Voice recogntion to ground
25. Play the recorded voice STOP
26. #if the alert message received
27. End the process
Fig.5 Hardware processing Flowchart

Fig. 4 Voice Detection Algorithm

978-1-7281-8501-9/21/$31.00 ©2021 IEEE 187

Authorized licensed use limited to: Dedan Kimathi University of Technology. Downloaded on May 18,2021 at 15:08:25 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the Sixth International Conference on Inventive Computation Technologies [ICICT 2021]
IEEE Xplore Part Number: CFP21F70-ART; ISBN: 978-1-7281-8501-9

Fig.4 explains algorithm and Fig.5 explains Fig 6 of bar chart. Added to that Sound
flowchar for voice detection. IOT cuts across various levels are shown This Graph represents that the
presentation domains ranging from civilian to sound can never be in a straight line graph. It changes
security sectors. Voice processing of instructions every second. From this data, can only deduce is the
received from the user starts with the node red skill peak value when there is a loud sound.
using the JavaScript, which debugs and deploy the Interconnection of uniquely identified computing
skill to the cloud. It tests with voice input whether all devices are defined by internet of things in the
the commands are recognized appropriately and existing infrastructure.
integrated with arduino UNO board. When it receives
API call from the cloud, arduino UNO understands 150
the command and checks the current status of the Decibel Scale
appliances and redirects message to API. arduino 100
UNO controller the instructions based on the received
message such as ON/OFF.
50
B. APPARATUS

The hardware requirement for this system are 0

 Arduino,
 Battery,
 LCD, Decibel Scale
 Button.

The Software requirements are Fig.7 Decibel scale

 Pick2kit and Fig.7 the future will be overlaid with

 Proteus. embedded devices of communication creating
“AUTOMATED LIBRARY SYSTEM”. A prototype
In this system have a very large library that will platform called arduino is used to (open source) give
difficult to manage. If the software is corrupted inputs to use software and hardware arduino uno
whole data will unsecure easily. It is very time board module-serial-voice recognition. Voice
consuming and it is extremely complex system. User commands can be stored by this module. Lot of
friendly environment and it is Dynamic and fully attention are gained by the internet and inertnet of
computerized system will develop with the data base things now. Sensory Substitution systems are very
of any data can be added, modified, deleted, very encouraging with the evolution of machine learning
quickly. High secure and scalable and reliable. It has and Personal computer technologies.
Record Maintenance.
IV. RESULT ANALYSIS

Fig. 6 Graph of Sound Record

Fig.8 Voice recognition status form

978-1-7281-8501-9/21/$31.00 ©2021 IEEE 188

Authorized licensed use limited to: Dedan Kimathi University of Technology. Downloaded on May 18,2021 at 15:08:25 UTC from IEEE Xplore. Restrictions apply.
 Proceedings of the Sixth International Conference on Inventive Computation Technologies [ICICT 2021]
IEEE Xplore Part Number: CFP21F70-ART; ISBN: 978-1-7281-8501-9

Fig. 8 form page represents the voice [3] Gino Lannace, Giuseppe Ciaburro, Amelia Trematerra, Heating
Ventilation and Air Conditioning (HVAC) Noise Detection in open
recognition status when the play button pressed from plan office using Recursive partitioning, Buildings 2018, 8(12),
the voice recognition board. 169.

A. SPECIFICATIONS [4] Maxence Guénin, Sophie Derelle, Marcel Caes, Laurent

Rubaldo , Isabelle Ribet-Mohamed, RT S Noise Detection and
Voltage Effect on RTS in HgCdT e Focal-Plane Arrays, Journal of
T ABLE III SPECIFICAT IONS OF ARDUINO
Electronic Materials,2020, volume 49, pages 6963–6970.

Operating the Voltage 5Volt [5] C. N. Vanitha; V. N. Jeevaa; S. Prajith Shriman,, Image and
for arduino. Face Recognition using CV Lens Machine Learning Android
Application, IEEE Xplore, 5th International Conference on
Advanced Computing & Communication Systems (ICACCS),
Recommended input 7 Volt to 12 Volt
2019.
voltage for the arduino.
[6] Zeenat Taria, Sayed Khushad Shah& Yugyung Lee, Smart 311
Limits for the input 6 Volt to 20 Volt Request system with automatic noise detection for safe
voltage. neighborhood, IEEE Xplore, IEEE International Smart Cities
Conference (ISC2), 2019.
Total input and output 14 pins labelled in form [7] Huanxian Bu,Wenjun Yu & Xun Huang, Wind-T unnel
pins. of (0-13). Investigation on the Compressive-Sensing T echnique for
Aeroengine Fan Noise Detection, AIAA Journal, v. 56, (9),
Total analog input pins 6 pins labelled in form September 2018, p. 3536-3546.
[8] C.Medina, R.Coelho, Impulsive Noise detection for
of (0-5) Intelligibility and Quality Improvement of speech Enhancement
Methods Applied in T ime Domain, Electrical Engineering and
Table III represents the specifications of the Systems Science, 2019.
arduino uno that are used in the
[9] QiWang,JingMa,SiyuanYu, LiyingT an Noise detection and
V.CONCLUSION image denoising based on fractional calculus,Elsevier Chaos,
Solitons & Fractals, Volume 131, February 2020, 109463.
The “AUTOMATION OF NOISE
DETECTION USING INTERNET THINGS” [10] N Sindhujaa, CN Vanitha, AS Subaira , An improved version
process made the surrounding to reduce the noise in of big data classification and clustering using graph search
technique, International Journal of Computer Science and Mobile
the way of controlling noise pollution. The main Computing, 2016, 5(2):224–229.
focus of this system is to reduce human voices.
[11] Mahmood Safaei, Abul Samad Ismail, Hassan Chizari, Maha
Several solutions have been proposed for identifying Driss, Wadii Boulila, Shahla Asadi, Mitra Safaei,Standalone noise
the voice detection. The maintenance of the noise is and anomaly detection in wireless sensor networks: A novel time‐
made efficient, as all the recorded voice are stored in series and adaptive Bayesian‐ network‐ based approach, 50(4),
2020, pp 428-446.
the secure database, through which data can be [12] Zuozhen Wang, Modified Rao test for distributed
retrieved easily. By this system, the person who target detection in interference and noise, Elsevier Signal
speaks in the enclosed area will get a mail to his/her Processing, 107578, Volume 172, July 2020.
mail address. This may ensures that the alert message
make the person who have been spoken and made the
person not to speak. Hence I conclude my words with
that the loud voice from a coworker may mentally
affect your mood. These system helps you from that
type of affect.

REFERENCES
[1] Sheng-Chieh Lee, Jhing – Fa Wan, G, Miao-Hia Chen,
T hreshold Based Noise Detection and Reduction for Automatic
Speech Recognition System in Human-Robot Interactions,
Sensors 2018, 18(7), 2068.
[2] T homaspadois, Olivierdoutres, Francksgard, Alainberry,
Optimization of a spherical microphone array geometry for
localizing acoustic sources using the generalized cross-correlation
technique, Elsevier Mechanical Systems and Signal Processing,
Volume 132, 1 October 2019, Pages 546-559.

978-1-7281-8501-9/21/$31.00 ©2021 IEEE 189

Authorized licensed use limited to: Dedan Kimathi University of Technology. Downloaded on May 18,2021 at 15:08:25 UTC from IEEE Xplore. Restrictions apply.