DESIGN AND DEVELOPMENT OF MUFFLER NOISE DETECTION DEVICE

A Project Development Study Presented to

The Faculty of College of Industrial Technology

BATANGAS STATE UNIVERSITY

Balayan, Batangas

In Partial Fulfillment

Of the Requirements for the Degree of

Bachelor of Industrial Technology

Major in Computer Technology

By:

Selwyn Rom D. Aguila

Maria Mholien R. Fruelda

Jenelyn D. Jayme

Jinno Anjello S. Maningat

 APPROVAL SHEET

This project development study entitled “DESIGN AND DEVELOPMENT OF

MUFFLER NOISE DETECTION DEVICE” prepared and submitted by Selwyn Rom D.
Aguila, Mholien Fruelda, Jenelyn Jayme, and Jinno Anjello S. Maningat, in partial
fulfillment of the requirements for the degree of Bachelor of Industrial Technology
Major in Computer Technology has been examined and recommended for acceptance
and approval for Oral Examination

CELINNE A. MENDEZ
Adviser

PANEL OF EXAMINERS
Approved by the committee on Oral Examination with the grade of _____

ENGR. RHOBERT E. ALVAREZ, DT

Chairman

ASST. PROF. MICHELLE M. DEL ROSARIO, LPT, PhD ENGR. ARMANDO E. ARZOBAL
Member Member

Accepted and approved in partial fulfillment of the requirements for the Degree of
bachelor of Industrial Technology Major in Computer Technology.

ENGR. RHOBERT E. ALVAREZ, DT

Date Campus Director/Head,

Academic Affairs

ii
 ACKNOWLEDGEMENT

The researchers are grateful to everyone who has supported and inspired this

study throughout its development and has contributed to its success.

First and foremost, to the almighty God, the source of all power, knowledge, and

capacities.

For years, they have used the Batangas State University Balayan - Campus as

their secondary home. The researchers are grateful for the opportunity to share their

knowledge and educate future scholars.

To the researchers' parents, relatives, and loved ones who have morally

supported them, made financial sacrifices for them, and given them hope that they may

achieve throughout their lives.

Celinne A. Mendez, for the encouragement, support, and guidance in conducting

the study, as well as for giving specific knowledge and recommendations that

contributed to the study's success.

ASRD

FMMR

JJD

MJAS

iii
 DEDICATION

This thesis is dedicated with genuine sincerity and deep affection to all those who

have provided unwavering support to the authors throughout their challenging journey.

We hold a special gratitude towards our parents, relatives, and loved ones, who have

not only supported us but also shown immense patience in dealing with our moments of

frustration and emotional breakdowns. Their words of encouragement have served as a

vital source of strength for us. We would also like to express our appreciation to our

friends, classmates, and teachers, whose assistance and motivation have been

invaluable throughout the entirety of this journey. Our heartfelt thanks go to our thesis

advisor and research instructor for their continuous guidance and support during the

entire process. And above all, we offer our utmost gratitude to the all-powerful God,

whose divine intervention has granted us the ability and resilience to successfully

complete this task. Oh Lord, we acknowledge and appreciate the abundant blessings

you have bestowed upon us. Praise be to God!

Selwyn Rom D. Aguila

Maria Mholien Fruelda

Jenelyn Jayme

Jinno Anjello S. Maningat

iv
 ABSTRACT

DESIGN AND DEVELOPMENT OF

TITLE :
MUFFLER NOISE DETECTION DEVICE

Selwyn Rom D. Aguila

AUTHORS :
Maria Mholien Fruelda
Jenelyn Jayme
Jinno Anjello Maningat

COURSE : Bachelor of Industrial Technology

MAJOR : Computer Technology

NAME AND
: Batangas State University, Balayan Campus
ADDRESS OF
INSTITUTION

ADVISER : Mrs. Celline A. Mendez

YEAR
COMPLETED : 2024

Summary

v
 As noise reduction continues to be a growing concern for the citizens of Balayan

Batangas, the study titled “Design And Development Of Muffler Noise Detection Device”

aims to create a systematic approach that would allow traffic enforcers to apprehend

violators of the Muffler Act of 2016, which aims to prohibit the use of modified mufflers

that would only increase the output of its volume, with much efficiency as these violators

tend to find openings or instances where they are able to showcase their mufflers while

also ensuring a way to escape from the area. The researchers’ intent is to be able to

design a device that would allow such action. Consequently, this type of design

becomes useful in other aspects other than the remote identification of any violators of

the Muffler Act of 2016. It also increases the level of security and safety of the road as

there is better surveillance around the area. Anyone that would opt to disregard proper

road discipline can be immediately apprehended. Additionally, the presence of this

device would be able to save the time of traffic enforcers and lessen their load of work.

From the pixels that appear on the screen, the road will be safer and more peaceful.

By using a combination of multiple devices, a device would be developed that

allows the operator to receive physical data and convert it into virtual information.

Thereby allowing for a more efficient approach in the problem. Usage of the Python

language for programming, all functions of the device can be expected to be in

accordance with its design, while also allowing for raw data to be interpreted into texts

of information accessible to the operator or manager. The system would also be able to

generate violator’s reports from the various data collected, providing their plate no.

vehicle profile, and the amount of output volume that their modified mufflers produce.

Multiple tests were conducted to ensure the functionality of the device and its accuracy.

vi
As it was within expectations, the device functioned as intended and provided concise

reports.

To recap, the muffler noise detection device that was designed by the team was

successfully crafted and has shown positive results with its function. The muffler noise

detection device underwent multiple tests which can be seen on figure 6.1 and 6.2. The

accuracy of the device can be observed in these figures as there is an indicator showing

its confidence level. It was also proven in this section how proximity has a relevant

influence on the accuracy of the device. Since the device is an early design of the

product, it is still under the assumption that it could only capture and identify vehicles

with a 50% level of confidence. However, that problem is accounted for by calibrating

the device to minimize the to be managed, thereby increasing the accuracy.

Findings

1. Relevant information can be gathered from the existing programs and

projects that tackle similar concepts like this. There are already existing

programs for road safety such as the CATH-ALL system, and it is these

kinds of programs that can provide the essental data needed to finish our

own design. Previous data showed the importance of their works thereby

further motivating the group to create a device that can help improve the

state of road safety and discipline.

2. By implementing flowcharts throughout the study, the researchers were able

to conclude the required steps to take to solve the complexity of the program

for the desired function. Especially for the software, a flowchart is essential in

creating an algorithm that follows a series of steps that allows for the

vii
 interpretation and analysis of data.

3. Upon the development of the software, the researchers concluded that

using Python for the overall programming code as it has a versatile

characteristic and is easy to use relative to other languages. Moreover, it

was also observed that Python as a programming language is better in

object-oriented programs.

4. The development stage of the project provided significant insights for the

effectiveness of the device. Although there are better options for certain

components such as the camera or the microphone, the affordability of

each were greatly considered during the gathering of materials.

Furthermore, the chosen components are guaranteed to be safe for use as

it possesses features that allow them to be safe even when experiencing

errors. Guaranteeing the function of the device was given priority over the

quality that may be gathered. In doing so, the economy for the project is

improved while also leaving room for future endeavors.

5. At a closer proximity of the subject, the accuracy level was able to reach up

to 90%, this the confidence level of the system with its reading. However,

as discussed during the early parts of the paper, at a longer distance, there

is a drop in its accuracy reading, making the previously 90% be 87.60%.

Considering that the space where the tests were held possesses an

enclosed structure, the evidence of the accuracy decreasing at the slight

difference in distance proves how the device is heavily affected by it.

Hence, will still be subjected under apprehensive inspection during its

viii
 operation to confirm the validity and credibility of the information provided

by the system.

Conclusions

The obtained results from the study have allowed the researchers to come

up with the following conclusions:

1. The final output done by the researchers is effective in terms of design,

operation, and safety.

2. It is concluded that flowcharts are as essential in planning as the materials

to be gathered. It provides a visual representation on the operation of the

project on both the hardware and software.

3. Through the use of proper programming language and coding, the

effectiveness and success of the device's function became apparent. It is

thus concluded that the following elements were essential in ensuring that

the project will work as designed.

4. The developed project, once implemented on the roads of Balayan

Batangas, specifically the locality of the Villa Crisanta Subdivision, it would

be able to aid the area in ensuring that violators of the law is properly

punished and apprehended. Consequently, it would also be able to

improve the overall safety of the road.

5. Determining the accuracy of the readings can be challenging for the

current setup but not impossible to solve. Due to the nature of the

components, proximity is the greatest adversary of the device as it hinders

ix
 its accuracy. Under certain circumstances, there is a chance that the

readings of the device may not be accurate or credible. But this should not

be taken as a flaw but as an opportunity for improvement.

6. The researchers concluded that the project is most effective when used in

subdivision streets or barangay streets where the road is not too busy in

order to ensure that the device functions as intended.

Recommendations

1. It is recommended that future studies should consider a longitudinal study

to determine the effectiveness of the project and the average number of

law violators as the safety and noise pollution surrounding an area may

vary in different scenarios.

2. This project itself possesses great potential for other fields it is

recommended that further endeavors for a similar study should include the

possibility of its application in different fields.

3. The researchers recommend the use of better components to collect better

samples of information. First is the usage of a camera with better video

quality (e.g. 1920x1080 high resolution camera), this would allow for better

view of the object. The placing of the amount or type of camera may also

be modified to manage the angle gaps existing within the frame. Lastly, the

microphone can be significantly improved by using a parabolic mic. Its

range and clarity are, by nature, much better than the currently available

for the group. Thus, it is recommended that it is used to gather better

x
 results.

4. Lastly, it is recommended that the programming for the device is enhanced

to create a more specific reading, thereby creating a more comprehensive

analysis on the vehicle’s violation. It can also enhance its tolerance level

for making errors in its readings.

TABLE OF CONTENTS

TITLE PAGE...................................................................................................................................i

APPROVAL SHEET......................................................................................................................ii

ACKNOWLEDGEMENT...............................................................................................................iii

DEDICATION...............................................................................................................................iv

ABSTRACT...................................................................................................................................v

TABLE OF CONTENTS..............................................................................................................xii

LIST OF TABLES.......................................................................................................................xiv

LIST OF FIGURES......................................................................................................................xv

CHAPTERS

I. THE PROBLEM AND ITS BACKGROUND

Introduction....................................................................................................1

Objectives of the Study.................................................................................2

Scope, Limitation, and Delimitation of the Study.......................................3

xi
 Significance of the Study..............................................................................4

Definition of Terms........................................................................................5

II. REVIEW OF RELATED LITERATURE

Conceptual Literature....................................................................................7

Research Literature.....................................................................................24

Synthesis......................................................................................................32

Conceptual Framework...............................................................................34

III. DEVELOPMENT METHODS AND PROCEDURES

Evaluation of Existing................................................................................. 36

Designing Stage...........................................................................................36

Tools and Materials Selection Stage..........................................................37

Development Stage......................................................................................37

Testing and Evaluation Stage.....................................................................38

Time Allotment.............................................................................................39

IV. PRESENTATION, ANALYSIS AND INTERPRETATION OF DATA

1. Evaluation of Existing.............................................................................41

2. Muffle Noise Detection Design...............................................................47

3. Tools and Materials Used........................................................................48

4. Development of Muffler Noise Detection Device.................................49

5. Testing and Evaluating the Project........................................................50

xii
CHAPTER V............................................................................................................... 57

Summary

Findings

Recommendations

LIST OF TABLES

TABLE NO. PAGE

1. GANTT CHART.................................................................................................39

xiii
 LIST OF FIGURES

FIGURE NO. PAGE

1. Conceptual Paradigm of the Study.....................................................................34

2. Operation Process of CATCH-ALL System........................................................46

3. System of Hardware and Software Design.........................................................48

4. Camera RTMP Server Setup................................................................................ 52

5. Sound Level Detection and Image to Data.........................................................54

6.1. Testing Accuracy...............................................................................................55

6.2. Variations in Accuracy......................................................................................55

xiv
 CHAPTER I

THE PROBLEM AND ITS BACKGROUND

This chapter includes the introduction, statement of the problem, objective,

scope, limitation and delimitation, significance of the study and definition of terms used

in the study.

Introduction

A muffler serves to dampen the noise produced by a vehicle's engine. Since the

engine has to produce a lot of power, it produces a lot of pulsating noise that

reverberates through the exhaust valves. These pulses are generated thousands of

times per minute. Without a properly functioning muffler, a vehicle can get incredibly

noisy. In fact, excessive noise is a good indicator that a new muffler is needed. Opening

the muffler reveals a series of pipes, baffles, and chambers. These components reflect

the sound waves produced by the engine. They are reflected so that the waves actually

cancel each other out. If the muffler is different, abnormal noise will also occur. Some

are designed to be silent, while others are specifically designed to have a growling

sound.

The Muffler Act of 2016 aims to prohibit motorbikes and all other types of

vehicles, both public and private, from employing modified mufflers that do nothing but

increase output volume. The bill also establishes a noise restriction for the cars,

mandating that they not exceed 70 decibels. Vehicles made for use in sporting events

and auto shows are exempt from the law. According to Louigi (2021), Ordinance No.
8722 or “Motor Vehicle Modified Muffler Noise Regulation Ordinance of the City of

Manila” is intended to
regulate and punish all types of motor vehicles with modified mufflers or tailpipes that

produce excessive and loud noise. It also points out that if the offender's vehicle fails

the noise level test twice, the municipality will make the necessary repairs and the

offender will pay for the repairs. If the perpetrator refuses to present the vehicle for

noise measurements, relinquish his driver's license, or otherwise defend himself against

confiscation of the vehicle, this is a willful violation of this rule and is presumed to be

one. In the municipality of Balayan, the ordinance is also being implemented. With the

above-mentioned ordinance in mind, the researchers come up with the idea of

conducting the designing and developing of a muffler noise detection device in order to

help the LTO and Police officials of Balayan to implement this ordinance. The device will

include a decibel meter for monitoring the sound of a vehicle passing by and a camera

that will be used to take photos of the vehicle’s plate number. The researchers will also

have a server where the data will be stored. With this study, the researchers aim to help

in implementing the ordinance No. 8722 in the municipality of Balayan, Batangas.

Objectives of the Study

The main objective of this study is to design and develop a muffler noise

detection device.

Especially this study aims to:

1. To evaluate the existing noise detection and image capturing:

1.1 Operation;

1.2 Design and;

2
 1.3 Safety

2. To design the muffler noise detection device in terms of:

2.1 Software Design and;

2.2 Hardware Design

3. To determine the tools and materials needed in developing the muffler noise

detection device

4. To develop the muffler noise detection device

5. To test and evaluate the development project in terms of:

5.1 Noise detection system;

5.2 Image capturing process;

5.3 Data storing process;

5.4 Data transfer;

5.5 Data extraction

Scope, Limitation, and Delimitation of the Study

The study aims to design and develop a muffler noise detection device. The study

does also evaluate the existing muffler detection device for the motorcycle vehicle. In

terms of design, safety and operation. Furthermore, the study also aims to design

devices that will detect the noise of mufflers of vehicles in terms of flowchart and

software design. In addition, to determine the tools and materials to be used. Lastly, this

3
study will test and evaluate the devices that may be used to detect the muffler noise in

forms of image processing accuracy, precision and accuracy of diagnosis.

The study is limited to motorcycle riders around the Villa Crisanta Subdivision at

the Municipality of Balayan. The study is also limited to different varieties of muffler.

This study is delimited for those vehicles that have more than three wheels.

However, the study excludes other subdivisions around Balayan except Villa Crisanta

Subdivision. Moreover, other sounds that are not detected by the decibel meter are also

excluded. The surveillance camera device does not consider recording video of vehicles

passing by.

Significance of the Study

This study will be beneficial to the Motorcycle Riders, Land Transportation Office,

and Future Researchers:

Motorcycle Riders. This proposed study will help them to comply with the Muffler

Act of 2016. This study will help them to be aware and to avoid violating the ordinance.

Land Transportation Office. This study aims to help in implementing the Muffler

Act of 2016 at the Municipality of Balayan.

Future Researchers. The concepts discussed in this study may serve as a basis

for future research or as a method of validating the accuracy of previous results.

Additionally, this can be used as a guide to give an overview of the design and

development of muffler noise detection devices.

4
Definition of Terms

For better understanding of the study, the following terms are defined

conceptually and operationally.

Decibel Meter. It's an electronic device used to gauge sound or noise levels, and

it's a tool enabling the assessment of the energy present in electromagnetic waves

(Kristjan Salev, 2021).

Muffler. The engine muffler is engineered with acoustic soundproofing in mind,

aiming to diminish the sound produced by the engine. Its design focuses on reducing

the intensity of sound pressure generated by the engine's operation (Shinde et al. 2017)

Data Extraction. Is the procedure of summarizing essential attributes of

research in a structured and standardized manner, relying on details found in journal

articles and reports (Lana Schmidt et al. 2021)

Database. A compiled assortment of organized data, designed for easy

management and updating. Information in the form of data records or files,

encompassing details like sales transactions, customer data, financial records, and

product specifics, is commonly gathered and housed within computerized databases.

Data transfer. Pertains to the movement of extensive datasets from one

organization or business division to another. This involves the secure exchange of files

between entities or systems.

Image processing. The process involves manipulating the image, which could

entail employing various methods until the desired outcome is achieved. The ultimate

5
result could manifest as an image itself or as a correlated image feature. Such an

outcome can be utilized for subsequent analysis and the process of making decisions.

6
 CHAPTER II

REVIEW OF RELATED LITERATURE

This chapter will present ideas, finished theses, generalizations or conclusions,

methodologies, syntheses, and conceptual frameworks with the main aim of getting

familiar with the revolving topic of design and development of muffler noise detection

devices. Such topics are relevant for this research.

Conceptual Literature

The researcher gathered different concepts in order to support the present study.

The concepts being discussed by this study are scanner, computer, programming

software and image processing used in the development of the muffler noise detection

device.

Software Development Life Cycle. According to Alexandra (2023), the

Software Development Life Cycle, often known as SDLC, is a method for creating

software that is of the greatest quality and least expensive in the shortest amount of

time. The well-organized phases of the SDLC allow an organization to swiftly develop

high-quality software that is well-tested and prepared for usage in production.

The introduction described the six phases of the SDLC. The waterfall model, spiral

model, and agile model are common SDLC models.

In accordance with Velimirovic (2022), The Software Development Life Cycle

(SDLC) is a comprehensive framework that defines the many steps of software

development, such as planning, developing, testing, and deploying. It is a systematic

guide to software development that assists teams in ensuring consistency, quality, and

efficiency throughout the process. An SDLC approach establishes a solid basis for
 software projects, allowing firms to fulfill deadlines, maintain code quality, avoid

problems, align features with business goals, and efficiently prioritize work. It entails

segmenting the development process into different stages, each having its own set of

goals, activities, and deliverables.

Based on Martyniuk (2022), SDLC (software development lifecycle) refers to the

process that a software project must go through and consists of a thorough plan that

describes how to build, maintain, replace, alter, or upgrade specific software. The

lifespan describes a way for enhancing software quality and the development process

as a whole. Analysis, Planning, Architecture Design, Development, Testing, and

Maintenance are the six stages of the SDLC.

Image processing. According to Kundu (2023), digital image processing

involves manipulating digital images using computer algorithms, making it an essential

preprocessing step in applications like face recognition, object detection, and image

compression. This process enhances existing images or extracts important information,

particularly in Deep Learning-based Computer Vision applications. Manipulating

images, such as adding or removing objects, is particularly important in the

entertainment industry.

As claimed by Hawley (2023), Image processing is a multidisciplinary field that

manipulates digital images using techniques and algorithms to enhance visual quality,

extract meaningful information, and enable automated analysis and interpretation. It has

applications in medical imaging, remote sensing, surveillance, entertainment, and

scientific research. Image processing revolutionizes how we perceive and interact with

visual content, impacting various industries and applications.

 In agreement with MacDonald (2013), image processing is a process that

involves translating between the human visual system and digital imaging devices. The

human visual system perceives the world differently than digital detectors, with display

devices imposing additional noise and bandwidth restrictions. To achieve translation,

image processing must be approached in a scientific manner, ensuring reproducibility

and validation of results. This involves recording and reporting processing actions, as

well as applying similar treatments to control images.

Programming language.

Database management system. According to Knight (2022), Database

Management is the process of organizing, storing, and retrieving data from a computer.

Database management may also refer to a database administrator's (DBA's) data

storage, operations, and security procedures throughout the data's life cycle. Designing,

implementing, and sustaining stored data to optimize its value is what database

management entails.

In agreement with Chia (2022), the process of organizing, storing, and modifying

data in an orderly fashion within a system is referred to as database management. This

management entails organized data handling from many applications and sources to

provide effective data validation, modification, and retrieval. Databases are data

repositories that can be conveniently accessed and manipulated to meet specific needs.

Database administration includes guaranteeing the data's quality and cleanliness, as

well as optimizing its storage for retrieval and modification operations.

23
 In line with BasuMallick (2022), a Database Management System (DBMS) is a

computerized solution that facilitates organized storage, manipulation, and retrieval of

information. It enables efficient data interaction, supports data analysis, and facilitates

data-driven workflows. DBMS categorizes data structures and types, modifies data

format, specifies validation rules, and ensures data integrity. It acts as an intermediary

between databases, end users, and application programs, maintaining organized and

accessible data through the database engine and schema. DBMS components ensure

data security, integrity, and concurrency, providing a centralized view for multiple users

and offering various levels of access control. It supports database design, connection

with software programs, data consistency, and security.

Research Literature

This Review of related literature is an important aspect for our investigation. This

will help the researchers to gather information about what is already been done in the

area of the study.

Based on Kundu and Hawley, image processing involves the manipulations of

digital images in order to enhance visual quality and extract valuable information. Image

processing constitutes a multidisciplinary domain centered around the manipulation of

digital images through techniques and applications such as face recognition, object

detection, and image compression. In an earlier perspective shared by MacDonald,

image processing emerges as a nuanced process bridging the gap between the human

visual system and digital imaging devices. Acknowledging the distinctive perspectives of

both, the translation between them involves intricate considerations. SDLC refers to the

24
process that a software project goes through, according to Alexandra, Velimirovic, and

Martyniuk. It is made up of plans that will allow a project to meet deadlines, maintain

quality, and minimize mistakes. The SDLC is divided into six steps, as outlined by

Alexandra and Martyniuk: analysis, planning, architecture design, development, testing,

and maintenance. According to Knight, Chia, and BasuMallick, Database Management

System is the process of organizing, storing and modifying data or information. Chia

defines it as systematic organization, storage, and modification of data within a system,

focusing on structured data handling. BasuMallick characterizes DBMS as a

computerized solution that streamlines data storage, manipulation, and retrieval,

supporting efficient data interaction, analysis, and workflows.

Based on the study of Philip Geneta (2019), the Arduino-based Decibel Meter for

Mufflers was developed to mobilize the roadside operations of the Land Transportation

Office and Local Government Units in controlling the noise pollution emitted by motor

vehicles with modified and defective mufflers. It was also developed to strengthen the

capacity of LTO and LGUxs in implementing the Muffler Act of 2016, an act prohibiting

the use of motor vehicles without mufflers or those that are defective or have been

modified which increase the sound emitted by motor vehicles, penalizing owners and

drivers that contribute to noise pollution in the environment. The researchers used the

developmental type of research which employed thorough planning and analysis to

achieve the objectives of the studies. The prototype has a mobile application that

enables the prototype to sense the sound level of the motor vehicle being tested,

capture the plate number, and locate the whereabouts of the vehicle in real-time and

send the report via SMS (Short Message Service). Thus, locally available materials

25
were used in the fabrication of the prototype such as Arduino Nano microcontroller,

sound sensor, serial camera, SD card module and HC-05 Bluetooth module. The mobile

application was developed with the software Basic4Andriod. The developed prototype

was tested for accuracy and functionality. Accuracy tests have shown that the sound

level readings of the prototype are 98% compared to those of sound level meters used

by licensed government agencies. Additionally, functional testing of the device showed

that the serial camera, SD card, and Bluetooth worked. Its performance was evaluated

by 10 of his IT professionals and programmers and his 10 potential end-users,

consisting of police officers, LTO personnel, and other law enforcement agencies. The

evaluations performed were based on profitability, functionality, quality, security, and

user interface design. The prototype received an average overall score of 4.35 and an

excellent descriptive rating. This means that the prototype served its purpose and the

evaluators were satisfied with the results and functionality of the Adruino-based muffler

decibel meter.

Geneta (2019) developed an arduino-based decibel meter for Mufflers to

mobilize the roadside operations of the Land Transportation Office and Local

Government Units and strengthen their capacity to implement the Muffler Act of 2016 by

controlling the noise pollution emitted by motor vehicles with modified and defective

mufflers. The researchers used the developmental type of research which employed

thorough planning and analysis to achieve the objectives of the studies. The prototype

has a mobile application that enables the prototype to sense the sound level of the

motor vehicle being tested, capture the plate number, and locate the whereabouts of the

vehicle in real-time and send the report via SMS (Short Message Service). Locally

26
available materials were used in the fabrication of the prototype such as Arduino Nano

microcontroller, sound sensor, serial camera, SD card module and HC-05 Bluetooth

module. The mobile application was developed with the software Basic4Andriod.

Results of evaluation showed that the sound level readings of the prototype are 98%

compared to those of sound level meters used by licensed government agencies.

Functional testing of the device revealed that the serial camera, SD card, and Bluetooth

worked efficiently. The study concluded that the prototype was accurate and functional

in detecting muffler noises.

According to Rodrigo Vázquez-López (2022), the health impact has increased

due to various environmental factors; among them, noise pollution due to high exposure

will cause diseases such as hearing, respiratory, cardiovascular, sleep disorders and

even headaches. This work proposes a monitoring system to measure noise, as well as

manage and store information obtained by the Internet of Things (IoT), to monitor the

behavior of this variable and be able to make decisions when it comes to noise. Sound

is expressed at excessive levels and it is harmful to human health and well-being. To

check the operation of the established monitoring system, two experiments were

performed: sensor characteristics are supported by instrumentation (sound level meter)

as well as sound level meter and system evaluation via IoT. As a result, noise is

removed from the input signal to obtain a smoothed output signal, then transmits and

stores the data in a database, and finally visualizes that data on a platform using the

Internet of Things model.

According to the study of Cagatay Catal (2011), noise detection for software

measurement datasets is a topic of increasing interest. The presence of noise class and

27
attribute in software measurement datasets degrades the performance of machine

learning-based classifiers, and identifying these noisy modules improves overall

performance. In this study, Catal proposes a noise detection algorithm based on the

threshold values of the software measure. The threshold values are obtained from the

analysis of the receiver operating characteristics (ROC). This article focuses on case

studies of five public NASA datasets and details the construction of software failure

prediction models based on Naive Bayes before and after applying the detection

algorithm. Suggested noise. Experimental results show that this noise detection method

is very effective in detecting class noise, and the performance of error predictors using

Naive Bayes algorithm with logNum filter will improve if the class labels of the models-

The identified noise module is corrected.

The study of Phong Le and Ivan Titov (2019) said, accurate Entity Linkers were

written for domains and languages where annotated data (that is, text associated with

knowledge bases) was available. However, little progress has been made in

environments with no or very limited amounts of labeled data (for example, law or most

scientific disciplines). This work shows how to learn to link mentions without using only

labeled examples, knowledge bases, and collections of uncommented text from

corresponding domains. To achieve this, the task is presented as a multi-instance

learning problem and relies on surface fitting to create the initial noise labels. Since the

training signal is weak and the replacement labels are noisy, the researchers introduce

a noise detection component to the model. This allows the model to recognize and

ignore noisy examples. Our method of collectively learning to detect noise and

28
associate entities far outperforms the surface matching baseline. It even approaches

supervised learning performance for a subset of entity categories.

In the study of Dukyong Yoon et al. (2019), biosignal data collected by patient

monitoring systems can provide important evidence for the detection or prediction of

important clinical events; however, the noise in this data interferes with its use. Since

deep learning algorithms can extract features without human annotation, this study

hypothesized that they could be used to screen electrocardiograms (ECGs) that could

not be detected. Acceptable noise includes. To test this, a deep learning-based model

for non-acceptable ECG screening was developed, and the results of this model's

screening were compared with medical expert interpretations. To develop and apply the

screening model, they used a biosignal database consisting of 165,142,920 ECG II data

(10 s ECG II data) collected from 31 May August 2016 to September 30, 2018 in an

intensive care unit of the trauma department. Then 2700 and 300 ECGs (ratio 9:1) was

reviewed and used by a medical professional for 9 cross-validation (training and

validation) and test datasets. A convolutional neural network-based model for non-

acceptable ECG screening was developed based on training and validation datasets.

The model with the lowest cross-validation loss is then selected as the final model. Its

performance is evaluated by comparison with a set of test data. When the proposed

model's screening results are compared with the experimental data set, the area under

the receiver operating characteristic curve and the model's F1 score is 0.93 and 0.80

(sensitivity = 0.88, specificity = 0.89, positive predictive value = 0.74 and negative

predictive value = 0.96).The deep learning-based model developed in this study can

effectively detect and filter unacceptable ECGs.

29
 According to Adriana Leal et al. (2018), the analysis and interpretation of heart

sounds (HS) can be severely hampered by noise pollution when the signal is received in

an uncontrolled environment. Then, signal processing methods are needed for robust

analysis of HS collected in different recording contexts. Some works have dealt with this

problem using complex calculations that often depend on precise segmentation of the

signal. Therefore, the objective of the present study was to develop a low-complexity

automated algorithm capable of discriminating between healthy and contaminated HS

signals (or cardiac recordings) recorded in different situations. Spectral features were

used to characterize the different behaviors of clean and noisy HS in the

electrocardiogram (PCG) recording under noisy conditions. In particular, besides the

usual noise associated with auscultation in uncontrolled environments, other noisy

sounds are deliberately simulated and include voice, ambient sounds, and other

physiological disturbances. Instead of HS. The available signals were recorded in 24

healthy volunteers and in 8 patients diagnosed with various cardiac disorders. Subjects

included in the healthy dataset followed a predefined protocol in which environmental,

physiological, and speech interferences were simulated. A total of 288 PCGs, recorded

at auscultation and mitral valve sites, comprised the healthy dataset. The pathology

data set consisted of 16 PCGs, obtained at lung and tricuspid locations. The described

method is compared with two existing methods developed for the same purpose. Our

algorithm was found to give the best performance on different types of noise, resulting

in average sensitivity and specificity of 88.4% and 85.6% for the healthy dataset,

respectively. Strong and 84.3% and 85.8% for the pathology data set. These results

correspond to an increase of up to 4.7% in SE and 39.0% in SP compared with the two

30
reference methods in a healthy data set. Regarding the pathology dataset, our method

improved noise detection by up to 27.0% in SE and 34.1% in SP.

As stated by the study of Borut Sluban et al. (2010), noise filtering is often used

in data preprocessing to improve the accuracy of inductive classifiers. Our goal is

different: aims to detect noisy cases to be examined by a domain expert during the data

discovery phase. Therefore, our noise detection algorithms must have superior noise

detection accuracy, where the accuracy recall trade-off is modeled using the F

measure. New variants of the algorithm A layered noise detection algorithm has been

developed, including a highly tuned random forest filter, which ensures very high

accuracy of the identified bad data cases.

The study of João Rodrigues et al. (2017) said that the Biological signals are

often contaminated with limb movement, muscle contraction, or electrical noise. Many

algorithms in the literature, such as thresholding and adaptive filtering methods, focus

on detecting these noise patterns. This study presents a novel method for detecting

noise and spuriousness in electrocardiograms based on time series clustering. The

algorithm starts by extracting the most characteristic features for the signal's shape and

behavior over time and grouping its samples into distinct clusters using agglomeration

clustering. This method has been tested in many datasets to show that it is independent

of specific records, and that overall, the algorithm can successfully detect noise patterns

and artifacts with a sensitivity of 88%, specificity of 92% and accuracy of 91%,

demonstrating good performance in pattern detection based on morphological

clustering. This algorithm can be applied to detect and classify many types of noise for

more accurate de-noising and suitable for signal classification.

31
 Based on the study of Luís P.F. Garcia et al. (2016), the presence of noise in real

data sets can harm the predictive performance of machine learning algorithms. There

are several noise filtering techniques whose goal is to improve the quality of the data in

classification tasks. These techniques usually scan the data for noise identification in a

preprocessing step. Nonetheless, this is a non-trivial task and some noisy data can

remain unidentified, while safe data can also be removed. The bias of each filtering

technique influences its performance on a particular data set. Therefore, there is no

single technique that can be considered the best for all domains or data distribution and

choosing a particular filter is not straightforward. Meta-learning has been largely used in

the last years to support the recommendation of the most suitable machine learning

algorithm(s) for a new data set. This paper presents a meta-learning recommendation

system able to predict the expected performance of noise filters in noisy data

identification tasks. For such, a meta-base is created, containing meta-features

extracted from several corrupted data sets along with the performance of some noise

filters when applied to these data sets. Next, regression models are induced from this

meta-base to predict the expected performance of the investigated filters in the

identification of noisy data. The experimental results show that meta-learning can

provide a good recommendation of the most promising filters to be applied to new

classification data sets.

According to the study of Borut Sluban et al. (2014), noise filtering is used most

often during data preprocessing to improve the accuracy of inductive classifiers. The

object of this work is different: aims to detect noisy cases to improve data

comprehension, data cleaning, and outlier identification. The article is made up of three

32
parts. The first part presents a set-based noise ranking method to unambiguously

identify noise and outliers, named NOISE-RANK, which has been successfully applied

to a real medical problem economy, as demonstrated by the evaluations of experts in

the field. The second part deals with the quantitative evaluation of the performance of

noise detection algorithms on data with randomly introduced noise. A method to

evaluate the visual performance of precision spatial noise detection algorithms, named

VIPER, is presented and compared with the standard evaluation method. The third

section presents the implementation of NOISERANK and VIPER methods in the web

platform for generating and implementing data mining processes. This implementation

allows for public accessibility of the developed methods, the ability to repeat and share

the tests presented, and the inclusion of web services that allow the integration of

detection algorithms. New noise in the proposed noise detection and noise detection

procedure.

Synthesis

Researchers used and collected other studies and research from the Internet,

unpublished papers, articles, etc. to determine the nature and relevance of this study.

Philip Geneta (2019), Similar to the proposed study, both studies aim to help

implement the Muffler Act of 2016. Geneta’s output includes sending an SMS of reports

unlike the proposed study which will have a server/portal that will display the reports.

33
 Rodrigo Vázquez-López (2022), she studied the health impacts of noise pollution

by performing experiments that involved monitoring noise which is similar to the present

study.

Cagatay Catal (2011), this study proposes a noise detection algorithm based on

the threshold values of the software measure, which are obtained from the analysis of

the receiver operating characteristics (ROC). Similar to the present study that will also

have a noise detection device.

Phong Le and Ivan Titov (2019), this work shows how to learn to link mentions

without using only labeled examples, knowledge bases, and collections of

uncommented text from corresponding domains. Just like the present study, both

studies developed a noise detection device.

Dukyong Yoon et al. (2019), their study developed a deep learning-based model

for non-acceptable ECG screening. Like the present study, both studies aimed to detect

noises. But unlike their study, the current study only focuses at detecting noises from

vehicle’s mufflers.

Adriana Leal et al. (2018), a study which analyses how noise pollution affects

heart sound. Similarly to the current study, both studies will be experimenting on

detecting noises.

Borut Sluban et al. (2010), their study aims to detect noise just like the current

study.

João Rodrigues et al. (2017), this study presents a novel method for detecting

noise and spuriousness in electrocardiograms based on time series clustering. Both

34
studies aim to detect noises but unlike the present study, their study focuses on

electrocardiograms.

Luís P.F. Garcia et al. (2016), presents a meta-learning recommendation system

able to predict the expected performance of noise filters in noisy data identification

tasks. But unlike the present study which will only focus on detecting noises from

mufflers.

Borut Sluban et al. (2014), their study aims to detect noisy cases to improve data

comprehension, data cleaning, and outlier identification which is similar to the present

study but different because the present study focuses on detecting muffler noises only.

Conceptual Framework

This part of the study shows the relationship of the input, process and the output

in developing the project.

Input Process Output

 Gathered data  Planning

and information
 Designing
 Technical Skills
 Appropriate  Developing Muffler Noise
programming Detection Device
 Testing and
language, tools
and equipment Evaluation

35
 Figure 1. Conceptual Paradigm of the Study

The first frame shows the inputs in implementing the project. This includes the

collection of data and information, as well as the ideas and concepts of researchers and

advisers. This includes researching important information for the development of the

project. This phase also includes the selection of suitable programming languages that

researchers can use in creating the detection device. These are the underlying inputs

for how researchers will create the project’s output.

The second framework is a process involving planning that helps to prepare this

research project. It also includes the design, creating the layout of the device and

inserting all the content required for the design. The development of the muffler noise

detection device is based on the scope of research and the functions/transactions that

the software can perform. Testing and evaluation are a necessary way to ensure and

verify that your project is responsive and running smoothly.

After the input and process stages, researchers proceeded to the third and final

stage, output. This phase enabled the final output which is the design and development

of the muffler noise detection device.

36
 CHAPTER III

DEVELOPMENT METHODS AND PROCEDURES

This chapter focuses on the techniques and procedure that the researcher

employed to accomplish the design and development of the muffler noise detection

device.

Evaluation of Existing

The researchers are evaluating the current device for motorcycles to detect the

noise of a muffler pipe in terms of design, operation, and safety. The researchers will

also visit or conduct site inspections at some crowded places where other motorcycles

pass by on this place or road and where our hardware devices will be placed.

Furthermore, from the evaluation of existing parts, the researchers will identify other

parts that need further enhancement.

Designing Stage

Moreover, the software is designed to come up with user’s deeper understanding

of how to design and develop the software and hardware that detect the muffler noise

detection device. Flowcharts are also provided to further explain the study’s procedure.

The Muffler noise detection device has a Data Base Management System

(DBMS is a software package designed to store, retrieve, query, and manage data. The

user interface (UI) allows authorized entities to create, read, update, and delete data.

DBMSs are important because they provide programmers, database administrators,

and end-users with a centralized view of data and free and non-invasive end-user
applications. Find out where the data is located. APIs (Application Program Interfaces)

manage requests and responses for specific types of data on the Internet, and also

Software Development Life Cycle (SDLC) provides the fastest possible production of

high-quality, low-cost software. Producing top-notch software that meets and surpasses

all client expectations and needs is the aim of the SDLC. The SDLC develops and

specifies a comprehensive plan with stages, or phases, each of which includes its own

procedure and outputs that can help the researcher make the software more reliable

and user friendly.

Tools and Materials Selection Stage

The design and evaluation of the existing study are considered a basis for

selecting the appropriate materials required to develop the proposed study.

Researchers determine the appropriate materials to purchase and use in the

construction of the project. Researchers will seek out the highest quality materials to

help achieve the goals of this study. Affordability and quality of materials are also

considered at this stage. Researchers will inquire about the price of the materials online

and in the local market. When materials are not available locally, researchers will rely

on online websites. Researchers will also choose the best and most suitable

programming language according to the requirements of the proposed study.

Development Stage

In developing the proposed study, researchers will consider the evaluation of

existing studies, the proposed design, and the materials selected to achieve the

planned desired results. When developing the project, the researchers consider the

37
data and information provided by the people consulted. Researchers also brainstormed

conceptual literature and existing research to acquire relevant knowledge and skills for

project development. When developing software, researchers also consider the use of

suitable and appropriate programming languages. Researchers will make sure that the

data and information obtained from the analysis of existing studies, the design and the

choice of materials contributed positively and significantly to the development of the

proposed study.

Testing and Evaluation Stage

After developing the noise detection device, the researcher will test and assess it

to make sure it can work as expected. This stage will evaluate how successfully the

project will work. Researchers will test and evaluate key components of the design and

development of a computer software, hardware that detect the muffler noise detection

device. To get reliable test results, the researcher will do five trials to detect the noise of

motors to test the accuracy of the output. Researchers will demonstrate and discuss the

project and its purpose to help users understand its functionality. Researchers use

muffler noise detection software and hardware to detect motorcycle noises to test if they

are working as intended. During testing and evaluation, researchers check whether the

software process is successful. After a series of tests and evaluations, the researchers

recorded the results and comments and analyzed the results of the tests and

evaluations. The researchers will also seek feedback from individuals and incorporate

their suggestions.

38
Time Allotment

The researchers will manage the time needed to accomplish the design and

development of Muffler Noise Detection Device.

Table 1.

Gantt Chart

PROCEDURE AUG SEP OCT NOV DEC JAN FEB MAR APR MAY

EVALUATING

STAGE

DESIGNING

AND

DEVELOPING

STAGE

SIMULATION

STAGE

TESTING

STAGE

Table 1 shows the completed phases of the study and the time allocated for each

phase. August through November will be spent evaluating the existing similar studies to

the current research. The collected data and information will be used to design the

muffler noise detector device. In December, researchers will begin the design and

39
development phase that determines the programming language and materials needed

to use in the development of the output. During the month of January, the researchers

test the output’s operative and functionality, which is accomplished until Ma

40
 CHAPTER IV

PRESENTATION, ANALYSIS AND INTERPRETATION OF DATA

In this chapter, the results of the design and development of Muffler Noise Detection

Device is presented and discussed.

1. Evaluation of Existing

International Journal of Computer and Electrical Engineering (2011), presents a

Speed Detection Camera System (SDCS) as an alternative to conventional radar

technology and is seen to be a good application for some challenging image

processing concepts and algorithms, such as object motion detection, shadow

removal, and object tracking.

1.1 Operation

The existing study which is the SDCS uses a number of image processing

techniques on a live video stream - captured from a single camera - or in offline

mode, which allows SDCS to calculate the speed of moving objects avoiding

problems. Subject of traditional radar. SDCS provides an expensive alternative to

traditional speed cameras with exactly the same or even better quality. SDCS

processes can be divided into four consecutive phases; the first step is the object

detection phase, which uses a hybrid algorithm based on a combination of the

adaptive background subtraction algorithm with three frames. The second step is

audience tracking, which includes three activities, audience segmentation, fugitive

tagging, and object extraction. Supports object tracking taking into account various
possible movement scenarios object like; In simple tracking, subject leaves the

scene, subject enters the

 scene, subject passes another subject and subject leaves, and another person

enters the scene. The third step is the speed calculation stage, calculated from the

number. The image is consumed by the subject to pass through the front of the

stage. The final is the subject photography phase, photographing the subject of a

speed violation. SDCS is implemented and tested in many experiments; turned out

to be satisfactorily implemented. Detecting moving objects in video streams is known

to be an important and difficult research problem. Individual the intrinsic utility of

being able to segment video Motion stream and background components, detect

moving points, provide a focal point for perception, categorize and analyze activities,

and return later more efficient processes because only pixels "move" is considered.

The purpose of object tracking is to establish a match between objects or parts of

objects in consecutive images. This one too for the purpose of extracting temporary

information about objects such as trajectory, posture, speed and direction. Tracking

detected objects frame by frame, in video is an important and difficult task. It is an

important part of the intelligent surveillance system. That is because the object is not

tracked, the system cannot extract consistent temporal information about objects. In

this case, higher level behavior analysis steps would not be possible. On the other

hand, an incorrect foreground object segment due to ball, reflection and tackle

makes it difficult to follow a research problem. Since the output of the object

detection phase is significantly reliable and it handles sudden light changes and

dark. Therefore, the foreground image is ready for segmentation, labeling and

tracking. In addition, the object's speed can be calculated by first frame detection

42
subject entering the scene at (Fr0) and following the object until it exits the frame

(Frn). That is while ignoring other unimportant objects like people.

1.2 Design

The novel SDCS theory and system are briefly discussed in this section. Precise

planning and project execution. The SDCS can be broken down into four

subsequent phases, including:

• Detecting objects

• Tracking of objects

• Calculating speed

• Taking a picture of an object

It is known that moving object detection in video streams can be a big and

challenging scientific issue. Apart from the value of being able to divide a video into

sections streams to identify moving and background elements Moving blobs give

recognition a center of attention, Analysis of activity and classification help create

these later procedures are more effective since only "moving" pixels must be

considered. Object tracking's purpose is to build a correlation. Between objects or

object components in subsequent frames. As well is intended to extract time

information from items like trajectories, stance, and speed. The purpose of object

tracking is to establish a match between objects or parts of objects in consecutive

frames. This one too for the purpose of extracting temporary information about

objects such as trajectory, posture, speed and direction. Tracking detected objects

frame by frame in video is an important and difficult task. It is an important part of the

intelligent surveillance system. That is because without tracking the object, the

43
system cannot extract Consistent temporal information about objects. In this case,

higher level behavior analysis steps would not be possible. On the other hand, an

incorrect foreground object Segmentation due to balls, reflections and tackles makes

tracking a research problem difficult. Since the output of the object detection phase

is significantly reliable and it handles sudden light changes and dark. Therefore, the

foreground image is ready for segmentation, labeling and tracking. In addition, the

object's velocity can be calculated by the first frame detection subject entering the

scene at (Fr0) and tracking the object until it leaves the frame (Frn). That is while

ignoring other unimportant objects like people crossing the road. After the first

iteration, obviously the researchers still need another iteration to give exact results,

besides there exist detected false regions that will disappear after an additional

repeat as shown below in figure (7) (a) (b) (c) Figs. (7). 2nd iteration object

segment.The second iteration separates the two objects at the same area shown in

Figure 7.a and false rejection detected regions. The third iteration is enough

because there are no more segments to do it, now comes another problem: how to

determine enough iterations to give accurate results? Simply the system performs

segmentation until the end segmentation is available, and therefore the method has

been proven flexible and adaptable to any particular situation Present. In the end,

this self-developed method turned out to be a reliable segmentation technique with

very impressive results and no need to ensure object connection before processing

with segmentation.

44
1.3 Safety

Multiple solutions have been thought out in order to enhance the efficiency of

traffic management and security around every existing road that frequently

experience heavy traffic. It is definitely quite a challenge to capture every road

violator out there, which is why the assistance of A.I. are incredibly helpful in fulfilling

the gaps of human capabilities. As of the current situation, more than 1.3 million

deaths are met during road accidents per year. It has become one of the leading

cause of deaths for all ages, which makes it more terrifying as it can happen to

anyone. Often, most accidents that happen on the road are likely due to violators

who cannot follow speed limits, proper maintenance, or road discipline. As a result,

government organizations such as the World Health Organization (WHO),

Department of Transportation (DOTr), and other road safety partners have launched

the Philippine Road Safety Action Plan 2023 – 2028 (World Health Organization, 2023).

Although it consists of five pillars that successfully represents the effective actions to

be taken regarding the improvement of road safety, most of the steps written in their

pillars involve taking care of the people, the road, and the vehicle. While it is

certainly essential, this action plan can be observed to be lacking any efficient way

to capture or identify any road safety violators regardless of any accident occurrence

or lack thereof. However, while it is not entirely presented within the action plan on

how they may solve the mentioned problem, previous projects have already been

done to answer that issue. By using AI-powered system for highways, the

Philippines can implement a type of system that enables enforcers to capture and

identify and violators without even being there. This AI system is known as the

45
 CATCH-ALL or Contactless Apprehension of Traffic Violators on a 24-Hour Basis

and All-Vehicle Detection System (Umali, 2019). It is an automated system that can

recognize plate numbers and vehicle profile on a real-time basis whenever it detects

it going beyond the speed limit or violating any traffic laws. As it captures the violator

with its camera, it transmits relevant information to the operators or managers which

contain the profile of the vehicle and of the owner.

Figure 2. Operation Process of CATCH-ALL System

As it can be seen on Fig.2, the camera would capture the scene on the road and

would transmit information to the software from its interpreted data to identify the

violation, the plate number, and the profile of the vehicle.

46
 The use of the system can be seen to have a positive towards our traffic

situations. Since its implementation, traffic violators have decreased in numbers. The

road has become safer for children as there are less chances for people to violate any

traffic law since there is a chance of immediate capture from the CATCH-ALL project.

2. Muffle Noise Detection Design

To create a device capable of detecting muffler noise while disregarding any

white noise or background noise, simply placing a camera and a microphone would not

provide proper solutions. A combination of software (programming) and hardware

(appropriate components) is required to achieve the desired goal. Our group did a lot of

research and studying to find the right structure that would create our system with

minimal error and to achieve high efficiency.

2.1 Software Design

Our team used Python for the programming language of the device as it is

simple to use and easier to understand. Moreover, it allows for better versatility

when it comes to creating appropriate functions for the required event. It is also

better for object-oriented programming, which is heavily required for the current

project since its purpose is to capture moving objects and identifying them to

record their actions. For the entire function of the project, multiple software

designs were used as it not only require the camera to identify a moving object, it

is also required from the system to create an output of said object in order to

have the manager or operator use the provided information. An interface is

created solely to have an administrator dashboard that would show all the

information gathered from the camera and microphone. It would real-time details

47
 such as the decibels, the strength of the sound, the accuracy of sound capture,

and the tolerance allowable to capture violator. For its output, a separate user

interface is created to store relevant data. From here, the date, violation, and

identification of the vehicle would be seen and can be printed if necessary.

2.2 Hardware Design

The key aspect of this project to ensure its success is the hardware

design. For this project an appropriate camera, microphone, microcomputer, and

other connecting wires are used to ensure the effectiveness of the device.

HARDWARE: SOFTWARE: FINAL OUTCOME:

CAPTURES AND ANALYZE AND READABLE

GATHERS INTERPRET RAW DATA INFORMATION OF
INFORMATION VIOLATORS (Plate no.,
violation type, etc.)

Figure 3. System of Hardware and Software Design

The figure shown above shows visual interpretation of how the hardware and the

software work collaboratively to attain the objective. The hardware is shown to be the

primary receiver of information which will then be delivered to the software for analysis

and interpretation. After all the necessary stages are done, essential information can be

extracted.

3. Tools and Materials Used

While there is a lack of singular structure for the device, the goal of the group is

to create a functioning prototype to achieve our goal. Which is why the output can be

48
seen as separate devices or individual components attached to each other.

Nonetheless, it is able to produce relevant results for our research. To capture the

visual evidence or have visual representations for data, a Closed-Circuit Television

(CCTV) camera is used as it is most fitting for its affordability and capability. This type of

camera is also good for operation during any time of the day since it has its own light

source for when night falls. It can also switch to night vision if the location is too dark for

any light to reflect on its lenses. For the microphone, it was quite a challenge to identify

the appropriate classification of microphone to use since there are variables needed to

consider such as the distance of sound to be captured and the exclusion of white noise

during recording. As such, shotgun microphone was chosen since its function is very

efficient and for its affordability. From an objective perspective, a parabolic mic would be

much better since it is able to capture sound from longer distances and could collect

better raw data for analyzation. However, as expected, this type of mic does not fit the

affordability range of the research. Nevertheless, a shotgun mic served as a good

substitute since it utilizes a condenser or foam that is able to disregard any white noise

within the area. Additionally, its structure is cylindrical with “gills” on its side of the front

part in order to block any sound coming from the left or right direction. It has two modes

which is the normal mode and the telescopic mode. While normal mode has a range of

1 – 2 meters and can hear sound from a cone-like range, the telescopic mode can

reach for 4 – 6 meters, however it is only able to record sounds in a straight manner,

making whatever is set in-front of it to be the only one subjected to sound recording.

The last major component of the hardware design is the microcomputer, also known as

49
the Raspberry Pi. It is a simple device that collects the data from the output and

converts it into readable information on the admin interface.

50
4. Development of Muffler Noise Detection Device

By amalgamating the individual components to form a mesh of collaborating

functions, we were able to finalize a system where our camera would react accordingly

towards vehicles that meet that goes above the allowable muffler noise. As the

components are highly functional as individual parts, the only step required to create our

output is to have all of them function in a manner that would allow the program to

interpret their collected data into essential information. By utilizing various cords and

connector, such as Microphone audio jack to USB, it is ensured that each major

component would be able to interact with one another and work synchronously without

any time-lag. Thus, making all information be available in real time. Aside from the

physical components that we needed for the device; an intangible component is also

used to ensure that all the data mixes together well. Using the programming language

Python to create a program or a software that would allow raw data to be interpreted

into certain values and become understandable for operators and managers.

5. Testing and Evaluating the Project

After the development of our software and hardware, the team underwent

multiple testing to check the functionality of the device and its accuracy. Throughout the

multiple tests done, positive results were collected in terms of:

51
5.1 Noise Detection System;

Early tests showed immediate variations on the sound received by the

device. As it is within the expectation, the strength of sound received by the

microphone is directly proportional to its distance, meaning that the further the

source of the sound is, the more difficult it is for the microphone to accurately

detect its decibel range. However, this can be addressed by modifying the mode

52
that the microphone is in. By switching it from normal mode to “telescopic” mode,

its range is highly enhanced at the sacrifice of being only able to capture

sound in a straight line, unable to capture sounds from the surrounding area.

To further delineate the difference between the two modes, the normal mode can

gather sound from its surroundings, however, this mode is only able to reach up

to 1 – 2 meters. Whereas the telescopic mode can gather sound data from 4

meters to 6 meters. It increases the range by 300% while also blocking sounds

coming from the direction of left and right, making whatever in front of it become

the main subject for the sound.

After few more testing to identify which mode would fit best to achieve

desired results, our team decided to use telescopic mode as it would allow us to

gather a more accurate data for decibels. The camera movement would then be

main object of dependence by the microphone as it would be moving in tandem

with the camera movement.

5.2 Image Capturing Process;

The image capturing relies on software programming to have an object

tracking feature built into it. This feature allows the device to detect the object

which violates the sound limit, giving time for the program to track the moving

object, calculate its speed, capture an image of the object, and provide essential

information to be interpreted by the identification program. Our tests prove to be

successful in this aspect as when an object passed by the camera, it began to

track its movement. Following its process, when the object made a sound that

exceeded the allowable threshold, it began to read the plate no. attached to it

53
and deliver the gathered data to the computer for interpretation. However, the plate

no. image that is captured is not directly and immediately identified by the

program. For this project’s prototype, the team decided to use an existing

program that contains an essential function which completes the device. The

existing program in question is known as the Plate recognizer, a website that

increases the accuracy of reading plate numbers. Yet, unless there is a better

camera that is able to deliver a high-definition video (1920x1080), the current version

of camera that the group uses still poses a 50-50 chance of accuracy or a 50%

confidence. This is concluded as our testing have shown results where the

written texts on the plate numbers of are blurry and pixelized. Additionally, the

angle of the camera poses itself to be a problem for future instances as there

may be situations where the vehicle may be moving too fast for the tracker to analyze

the contents.

54
 Figure 4. Camera RTMP Server Setup

Fig. 4 shows a step-by-step process on how to properly set up the camera that

would be used for the device.

5.3 Data Storing Process;

Data collected during the detection processes is then stored at the

microcomputer that is connected to the same wi-fi as the noise detection device.

To set up the component, an ethernet cable is to be used along with the camera

to ensure that they are connected and can be accessible if the operator is

within the vicinity. Any data gathered by both the camera and the microphone will

be stored and processed within this small device.

5.4 Data Transfer and;

Every information gathered by the camera and the microphone is

interpreted through the microcomputer and inputted into the coding program

designed by the group in order to create a translate the raw data into readable

information. The stored data is transferred to the software which would then

transfer the information into the local website. Within this local website, various

options can be seen for any action that the operator wants to do. It contains

information from different days of when the device was operational, this is to

show the violators’ report. It also allows the user to choose a specific date or footage to

gather information from.

5.5 Data Extraction

55
 Data extraction can be found on the local website violator’s report tab.

From the selection of days there, where the violator of each day is listed, the user

can click on any of the selection and detailed information would be shown on the

screen. It would then be visible to the viewers the type of violation done by the

vehicle, its plate no. and the day and time it was taken. The chosen data can be

exported as a pdf file for a better formatting of information, or it can be printed

which shows all the essential information gathered and interpreted by the

program.

Figure 5. Sound Level Detection and Image to Data

Figure 5 provides a comprehensible guide on the function of the device’s system.

It shows the following steps that in undergoes starting from its sound detection. After the

recording of the data collected from its detection, it would be then applied to the

56
software program where the data will be extracted to be restructured into a systematic

information usable for the operator.

57
 Figure 6.1. Testing Accuracy

Figure 6.2. Variations in Accuracy

The two figures shown above, provide information regarding the accuracy of the

reading of the device. These data prove how distance can affect the performance of the

device as it hinders the rate of which sound can be detected. As it stands, every aspect

of the device seems to be functional. The violation report shows the various information

that it was designed to collect. It is able to read the plate number of the vehicle, identify

the date that it was captured, and provide a reading on the range of decibels that the

muffler was producing.

58
 CHAPTER V

SUMMARY, FINDINGS, CONCLUSIONS AND RECOMMENDATION

This chapter provides the summary of the paper, an in-depth analysis of its findings, and

various recommendations that can be taken as valuable information to consider for further

research to be done on this subject.

Summary

As noise reduction continues to be a growing concern for the citizens of Balayan

Batangas, the study titled “Design And Development Of Muffler Noise Detection Device”

aims to create a systematic approach that would allow traffic enforcers to apprehend

violators of the Muffler Act of 2016, which aims to prohibit the use of modified mufflers

that would only increase the output of its volume, with much efficiency as these violators

tend to find openings or instances where they are able to showcase their mufflers while

also ensuring a way to escape from the area. The researchers’ intent is to be able to

design a device that would allow such action. Consequently, this type of design

becomes useful in other aspects other than the remote identification of any violators of

the Muffler Act of 2016. It also increases the level of security and safety of the road as

there is better surveillance around the area. Anyone that would opt to disregard proper

road discipline can be immediately apprehended. Additionally, the presence of this

device would be able to save the time of traffic enforcers and lessen their load of work.

From the pixels that

appear on the screen, the road will be safer and more peaceful.

By using a combination of multiple devices, a device would be developed that

allows the operator to receive physical data and convert it into virtual information.

Thereby allowing for a more efficient approach in the problem. Usage of the Python

language for programming, all functions of the device can be expected to be in

accordance with its design, while also allowing for raw data to be interpreted into texts

of information accessible to the operator or manager. The system would also be able to

generate violator’s reports from the various data collected, providing their plate no.

vehicle profile, and the amount of output volume that their modified mufflers produce.

Multiple tests were conducted to ensure the functionality of the device and its accuracy.

As it was within expectations, the device functioned as intended and provided concise

reports.

To recap, the muffler noise detection device that was designed by the team was

successfully crafted and has shown positive results with its function. The muffler noise

detection device underwent multiple tests which can be seen on figure 6.1 and 6.2. The

accuracy of the device can be observed in these figures as there is an indicator showing

its confidence level. It was also proven in this section how proximity has a relevant

influence on the accuracy of the device. Since the device is an early design of the

product, it is still under the assumption that it could only capture and identify vehicles

with a 50% level of confidence. However, that problem is accounted for by calibrating

the device to minimize the to be managed, thereby increasing the accuracy.

Findings

58
1. Relevant information can be gathered from the existing programs and

projects that tackle similar concepts like this. There are already existing

programs for road safety such as the CATH-ALL system, and it is these

kinds of programs that can provide the essental data needed to finish our

own design. Previous data showed the importance of their works thereby

further motivating the group to create a device that can help improve the

state of road safety and discipline.

2. By implementing flowcharts throughout the study, the researchers were able

to conclude the required steps to take to solve the complexity of the

program for the desired function. Especially for the software, a flowchart is

essential in creating an algorithm that follows a series of steps that allows

for the interpretation and analysis of data.

3. Upon the development of the software, the researchers concluded that

using Python for the overall programming code as it has a versatile

characteristic and is easy to use relative to other languages. Moreover, it

was also observed that Python as a programming language is better in

object-oriented programs.

4. The development stage of the project provided significant insights for the

effectiveness of the device. Although there are better options for certain

components such as the camera or the microphone, the affordability of

each were greatly considered during the gathering of materials.

Furthermore, the chosen components are guaranteed to be safe for use

as it possesses features that allow them to be safe even when

59
 experiencing errors. Guaranteeing the function of the device was given

priority over the quality that may be gathered. In doing so, the economy

for the project is improved while also leaving room for future endeavors.

5. At a closer proximity of the subject, the accuracy level was able to reach up

to 90%, this the confidence level of the system with its reading. However,

as discussed during the early parts of the paper, at a longer distance,

there is a drop in its accuracy reading, making the previously 90% be

87.60%. Considering that the space where the tests were held possesses

an enclosed structure, the evidence of the accuracy decreasing at the

slight difference in distance proves how the device is heavily affected by

it. Hence, will still be subjected under apprehensive inspection during its

operation to confirm the validity and credibility of the information provided

by the system.

Conclusions

The obtained results from the study have allowed the researchers to come

up with the following conclusions:

1. The final output done by the researchers is effective in terms of design,

operation, and safety.

2. It is concluded that flowcharts are as essential in planning as the materials

to be gathered. It provides a visual representation on the operation of the

project on both the hardware and software.

3. Through the use of proper programming language and coding, the

60
 effectiveness and success of the device's function became apparent. It is

thus concluded that the following elements were essential in ensuring that

the project will work as designed.

4. The developed project, once implemented on the roads of Balayan

Batangas, specifically the locality of the Villa Crisanta Subdivision, it would

be able to aid the area in ensuring that violators of the law is properly

punished and apprehended. Consequently, it would also be able to

improve the overall safety of the road.

5. Determining the accuracy of the readings can be challenging for the

current setup but not impossible to solve. Due to the nature of the

components, proximity is the greatest adversary of the device as it hinders

its accuracy. Under certain circumstances, there is a chance that the

readings of the device may not be accurate or credible. But this should not

be taken as a flaw but as an opportunity for improvement.

6. The researchers concluded that the project is most effective when used in

subdivision streets or barangay streets where the road is not too busy in

order to ensure that the device functions as intended.

Recommendations

Introductory statement here

1. It is recommended that future studies should consider a longitudinal study

to determine the effectiveness of the project and the average number of

law violators as the safety and noise pollution surrounding an area may

61
 vary in different scenarios.

2. This project itself possesses great potential for other fields it is

recommended that further endeavors for a similar study should include the

possibility of its application in different fields.

3. The researchers recommend the use of better components to collect better

samples of information. First is the usage of a camera with better video

quality (e.g. 1920x1080 high resolution camera), this would allow for better

view of the object. The placing of the amount or type of camera may also

be modified to manage the angle gaps existing within the frame. Lastly, the

microphone can be significantly improved by using a parabolic mic. Its

range and clarity are, by nature, much better than the currently available

for the group. Thus, it is recommended that it is used to gather better

results.

4. Lastly, it is recommended that the programming for the device is enhanced

to create a more specific reading, thereby creating a more comprehensive

analysis on the vehicle’s violation. It can also enhance its tolerance level

for making errors in its readings.

62