Republic of the Philippines

Department of Education
Tacurong National High School
SCIENCE, TECHNOLOGY and ENGINEERING
PROGRAM
New Isabela, Tacurong City

CONTINUOUS OBSERVATION AND UTILIZATION OF

NUMERICAL TRACKING
(COUNT) SYSTEM

A Research Requirement
Presented to
The Department of Science,
Science, Technology, and Engineering Program,
Tacurong National High School, Tacurong City

In Partial Fulfillment
Of the Requirements in Research IV

By

NATHANIEL C. GAURANA
ALTHEA AYER S. JAPONES
MAUI KARYLL J. ESTANISLAO
HELAENA NICOLE B. DELACRUZ

APRIL 2024
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

ABSTRACT

Reliable people counting and person recognition systems are becoming more and

more necessary in today's quickly changing technology landscape, especially in

dynamic situations like train stations and retail malls. Current solutions frequently

have drawbacks that make it difficult for them to be used in busy interior locations. In

order to meet this urgent demand, the researchers have created a human counting

machine that is intended to be used in public spaces. This creative solution uses

technology like computer vision and machine learning to deliver precise and real-

time insights into visitor counts. User trials and surveys were used to collect data,

and descriptive statistics were used for analysis. The results validate the system's

functional performance by showing constant response times and great accuracy in

recognizing human presence. Furthermore, survey results show excellent

performance in overall functionality and data dependability. The user interface

should be optimized, compatibility issues should be resolved, and long-term

reliability tests should be carried out. All things considered, this research advances

the creation of precise and effective human counting systems that have numerous

uses in public safety, transportation planning, and urban planning.

ii
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

INTRODUCTION

Background of the Study

Reliable people counting and human detection are critical challenges in the field of

visual surveillance. While significant progress has been made in recent years, existing

solutions often come with limitations. These limitations include the necessity for people to be

in motion, the requirement for simple backgrounds, and the dependency on high-resolution

images (Hou & Pang, 2010, as cited in Raghavachari, Aparna, Chithira, & Balasubramanian,

2015). Such constraints hinder the applicability of these solutions across various indoor

scenarios, particularly in environments like malls where foot traffic can be complex and

dynamic.

People counting holds immense importance across numerous sectors, including urban

planning, transportation management, and public health. Manual counting methods are often

inadequate, especially in bustling environments like malls. Hence, there's a growing demand

for automated systems capable of providing real-time and accurate insights into visitor

numbers.

Existing research has highlighted the need for evaluating vision-based people

counting algorithms across different scenarios, environmental conditions, and camera

orientations to determine the most suitable solution for deployment (Raghavachari et al.,

2015). This underscores the necessity for specialized solutions tailored to specific

environments, such as malls.

The study aims to fill this gap by developing a Human Counting Machine, specifically

designed for deployment in public areas such as malls. This innovative system will leverage

advanced technologies, including computer vision, machine learning, and sensor integration,

1
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

to achieve a high level of accuracy in counting and monitoring human populations in real-

time.

Statement of the Problem

Generally, the study aimed to determine the potentiality of Continuous

Observation and Utilization of Numerical Tracking (COUNT) System.

Specifically, it answered the following questions:

1. What are the system capabilities of the COUNT System, as perceived by users,

with respect to:

a) Overall functionality

b) User friendliness

c) Performance in terms of speed and responsiveness of web

d) Data reliability

2. What is the functionality of the COUNT System, in terms of:

a) accuracy and error rate in detecting human presence

b) Time responsiveness

Hypotheses

HO1: Users perceive the overall functionality of the COUNT System positively.

2
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

HO2: Users find the interface of the COUNT System to be user-friendly and easy to
navigate.

HO3: Users perceive the COUNT System to have satisfactory speed and
responsiveness.

HO4: Users believe that the COUNT System has adequate security measures in
place.

HO5: Users perceive the data provided by the COUNT System to be accurate and
precise.

Significance of the Study

The significance of this study lies in designing an accurate human counting

machine with a database system, offering real-time data for improved foot

monitoring, resource planning, public safety, and efficient management across

diverse sectors.

The following will benefit from this study:

The business and retail sectors. Retail organizations can gain from the use of a

human counting machine with a database system by automating consumer counting,

assisting with inventory management, and providing useful information for marketing

campaigns.

Traffic Control and Transportation. By counting and analyzing human traffic at

different sites, cities and transportation authorities can improve overall urban

planning, manage traffic flow, and maximize public transportation routes.

3
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

shopping centers. Shopping centers can employ technology, much like the retail

industry, to track and analyze foot movement from customers.

Conceptual Framework

Figure 1. Conceptual Framework of the study

The conceptual framework of the study revolves around the integration of

various system device components, including E18 Infrared sensors, motion sensors,

electronic components, and software development tools such as Sublime Text web

app editor. These components are utilized for data gathering and analysis, focusing

on continuous observation and numerical tracking (count) system. Through the

development and capability enhancement of software, the system aims to achieve

efficient data analysis, enabling real-time monitoring and utilization of the gathered

information. This framework emphasizes the synergy between hardware and

software elements to create a system capable of accurate and timely data

processing for effective decision-making and action

4
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

Scope and Limitations of the Study

This study was limited only on determining Continuous Observation

and Utilization of Numerical Tracking (COUNT) System in terms of web and

system functionality Anything outside the bonds of the study will not be tackled

Time and Place of the Study

The study was conducted on January 22, 2024 – March 27, 2024. Brgy.

Sanpablo, Tacurong City, and Brgy. New Isabela, Tacurong City, Sultan Kudarat,

Philippines.

Definition of terms

Accuracy. The degree of correctness or precision in the measurements or data

obtained from a system, device, or process.

COUNT System. An acronym for Continuous Observation and Utilization of

Numerical Tracking System, referring to the specific system or technology being

studied.

Error. Any deviation or discrepancy between the observed or measured value and

the true or expected value, indicating inaccuracies or flaws in a system or process.

Precision. The level of exactness or consistency in the measurements or data

obtained from a system, device, or process.

5
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

Response Time. The interval between the initiation of a request or stimulus and the

corresponding response or action from a system or device.

REVIEW OF RELATED LITERATURE

Quan et al. (2023) introduced an innovative solution aimed at enhancing real-time

human detection and counting specifically tailored for the dynamic environment of

shopping malls. Their system, equipped with a user-friendly graphical interface and

comprehensive management functionalities, sought to address the challenges

associated with accurately monitoring foot traffic in crowded retail spaces. Through

an extensive review and comparative analysis of existing techniques and similar

systems, the study meticulously identified and evaluated various approaches to

select the most suitable solution tailored to the unique requirements of this scenario.

In a study by Kohli et al. (2020), the focus shifted towards leveraging machine

learning algorithms for sales forecasting, a pivotal aspect contributing to a company's

overall success. Emphasizing the critical role of accurate sales prediction models in

mitigating risks and facilitating informed decision-making, the researchers employed

advanced statistical methods such as Root Mean Square Error (RMSE) and Mean

Absolute Percentage Error (MAPE) to rigorously evaluate the performance of their

predictive models. By harnessing the power of machine learning, the study aimed to

provide businesses with actionable insights to optimize inventory management and

revenue generation strategies.

6
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

Addressing the challenge of reliable and real-time people counting, particularly in

densely populated environments, Arpit et al. (2021) proposed a comprehensive

methodology encompassing innovative techniques such as dimensionality reduction,

two-stage cascade-of-rejectors method, and fusion approaches for handling

occlusions in multi-camera setups. Through a series of meticulously designed

experiments, the researchers demonstrated the superiority of their approach over

existing state-of-the-art techniques, offering enhanced accuracy and efficiency in

people counting applications across diverse scenarios ranging from retail stores to

transportation hubs.

In a study by Lalchandani et al. (2021), the significance of people counting in various

commercial settings was underscored, highlighting its relevance for optimizing

operational efficiency and resource allocation. By illuminating the multifaceted utility

of people counting data for shop owners, security officials, transportation operators,

and disaster management agencies, the researchers emphasized the broader

implications of their findings for enhancing situational awareness and decision-

making capabilities across different domains.

Finally, Ikemura et al. (2011) proposed an innovative approach to human detection

utilizing Relational Depth Similarity Features (RDSF) extracted from Time-of-Flight

(TOF) camera data. Their method, characterized by high detection rates and low

false positive rates, offered real-time processing capabilities suitable for applications

requiring precise and efficient people counting. By harnessing the depth information

captured by TOF cameras, the researchers demonstrated the potential of their

7
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

approach to address the challenges associated with accurate and reliable human

detection in diverse real-world environments.

MATERIALS AND METHODS

Materials

The materials used in the study were the following: Arduino Uno, E-18D80NK

Infrared Obstacle Avoidance Photoelectric Sensor, jumper wires, 20x4 LCD Display,

Piezzo buzzer, breadboard, 3x3 feet plywood, 5 feet metal pedestal stand, Mounting

bracket , screw, and Cardboard.

Methods

Acquisition of Materials

Before developing the COUNT System, careful consideration was given to acquiring

essential materials. These materials, encompassing both hardware components and

web-based technologies, were selected to enable the system's functionality in

detecting human presence, tracking movements, and facilitating data storage and

analysis via a web platform.

Sensor/s:

 Infrared motion sensor (E-18D80NK): Quantity: 2

Microcontroller/s:

 Microcontroller (Arduino Uno): Quantity: 1

8
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

Software Tool/s:

 Sublime Text

 Xampp

- Used for coding web-based technologies

Prototyping Component/s

 Breadboard: Quantity : 1

 Mounting Brackets: 2

Structural Material/s

 Plywood: (3x3)

 Metal pedestal stand: (5 feet)

 Cardboard: 2x2

Electronic Component/s

 Jumper wires: Quantity: Bulk/Large Quantity

 20x4 LCD Display: Quantity: 1

 Piezzo buzzer: Quantity: 1

Tools:

 Mounting hardware: Brackets and screws

 Wiring accessories: soldering iron, screw driver/s

System Integration and Setup

9
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

Programming and setting up the circuits for COUNT System. This involves writing

and refining the instructions that control the device's operations, connecting the

hardware components together to ensure they function correctly, and ensuring that

the system operates smoothly and functions as intended.

Figure 2. System Component Connectivity Diagram

Programming the Infrared motion sensors

Programming the the Infrared motion sensors (E-18D80NK). This proces involves

setting up the code for the sensors to be able to work/execute and count human

entries accurately.

10
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

Figure 3. Infrared motion sensor setup

Interface Development for web

Crafting the Web Interface. This entails creating an overarching design, refining the

interface, and constructing the entire web platform. This encompasses developing

the website and implementing specific features, as well as organizing the web code

to enhance user-friendliness.

Figure 4. User Interface Design for the COUNT System Web Platform

11
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

Construction of Interface for system

Crafting the device Interface, creating a comprehensive design, and crafting specific

friendly design for the COUNT device. This process includes modeling the device,

cutting the cardboard, shaping the plywood, and starting to construct to get the

expected outcome of the device.

Figure 5. Construction of Device Interface for the COUNT System

Data Gathering Procedures

The study aimed to assess users' perceptions and experiences with the

COUNT System according to predefined objectives. Data were gathered through a

user survey. A 5-point scale was employed in the survey questionnaire to assess

various aspects of the COUNT System's capabilities, functionality, usability, and

performance. The scale ranged from 1 to 5, with 1 representing "Very Dissatisfied"

and 5 representing "Very Satisfied". Participants were asked to rate their level of

satisfaction or agreement with specific statements or attributes related to the COUNT

12
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

System. While a stopwatch will be used to measure the response time. Additionally,

to test the accuracy of the system device in counting human presence, data will be

collected through trials.

Statistical Analysis Tool

The statistical tool employed for the analysis of the study is descriptive

statistics. To assess the web-user functionality, performance, friendliness, and

responsiveness, descriptive statistics were utilized by calculating the means of the

ratings provided by the 30 respondents. In contrast, for response time and accuracy,

the data will be analyzed using the mean.

RESULTS AND DISCUSSIONS

Table 1. COUNT System’s Functionality in terms of Time Response in

Detecting Human Entries

TRIALS RESPONSE TIME (ms)

1 12
2 14

3 11
4 13
5 10
6 12
7 12
8 10
9 11
10 14

13
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

The table 1 presents response time measurements for ten trials of the

COUNT System device. Response times are recorded in milliseconds (ms). Trial 5

exhibited the shortest response time of 10 ms, while trial 7 demonstrated the longest

response time of 15 ms. The average response time across all trials is approximately

12.5 ms. These results suggest that the COUNT System performs consistently with

rapid response times, indicating functional performance.

Table 2. COUNT System’s Functionality in terms of Accuracy and Error rate in

Detecting Human Entries
TRIALS DISCREPANCY SYSTEM ACCURACY

1 NO ERROR PRECISE
2 NO ERROR PRECISE
3 NO ERROR PRECISE

4 NO ERROR PRECISE

5 NO ERROR PRECISE
6 NO ERROR PRECISE

7 NO ERROR PRECISE

8 NO ERROR PRECISE

9 NO ERROR PRECISE

10 NO ERROR PRECISE

The table 2 shows the:

Error: Throughout all ten trials, no errors were encountered, indicating the reliability

and robustness of the system in accurately detecting human presence

14
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

Detection Accuracy: In each trial, the data recorded by the system was deemed

accurate, demonstrating the system's effectiveness and accuracy in determining

human presence.

Based on the data presented, it can be concluded that the COUNT System

functions effectively in terms of both error prevention and detecting human presence.

Table 3. COUNT System’s Web Survey result; using a 5 point scale

EVALUATION TOTAL

AVERAGE

Overall -- -- 4.11

funtionality

User friendliness -- -- 3.53

System -- -- 4.00

responsiveness

Data Reliability -- -- 4.31

Total -- -- 3.98

The table 3 shows the survey findings for the performance evaluation of the

COUNT System are summarized in Table 6, detailing average scores across

different aspects.

15
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

The system demonstrates commendable performance in several areas. Notably,

users express high satisfaction with the system's overall functionality, reflected in an

average score of 4.11. This underscores the system's effectiveness in meeting user

needs and delivering desired outcomes.

Nevertheless, the system earns praise for its responsiveness, scoring an average of

4.00, and data reliability, with the highest average score of 4.31. These aspects are

crucial for ensuring efficient user interactions and maintaining trust in the accuracy of

data generated by the system.

In summary, while the COUNT System demonstrates strengths in functionality and

reliability, there is room for improvement in user experience and compatibility.

CONCLUSION AND RECOMMENDATIONS

Conclusion

Based on the data gathered during the conduct of the study;

In response time measurements for the COUNT System device across ten trials. The

data showcases consistent and rapid response times, with an average of

approximately 12.5 milliseconds. This indicates that the system operates efficiently,

demonstrating functional performance in detecting human presence.

In the absence of errors and high detection accuracy across all ten trials

underscores the reliability and effectiveness of the COUNT System in accurately

detecting human presence. This highlights the system's robustness in error

16
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

prevention and its capability to accurately determine human presence, further

affirming its functional performance.

According to the survey findings for the performance evaluation of the COUNT

System. The system demonstrates commendable performance in overall

functionality, system responsiveness, and data reliability, as reflected in the high

average scores. While user friendliness could be improved, particularly in enhancing

the user experience and compatibility, the system's overall effectiveness and

reliability are evident.

In conclusion, the COUNT System showcases strengths in functionality, reliability,

and accuracy. While there are areas for improvement identified, such as user

friendliness, the system's overall performance is commendable. Addressing user

feedback and optimizing system controls can further enhance the system's usability

and effectiveness, ensuring continued satisfaction and trust among users.

Recommendations

To improve the functionality and usability of the COUNT System based on our

study findings, the following recommendations are suggested:

1. gather detailed feedback to streamline the user interface and improve overall
usability.

2. Optimize Cross-Device Compatibility: Address challenges in cross-device

compatibility through compatibility testing and ensure consistent user experience
across different devices.

17
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

3. Long-Term Reliability Assessment: Conduct longitudinal studies to evaluate the

system's performance and accuracy over extended periods, ensuring durability and
stability.

4. Scalability and Integration: Explore opportunities to scale up the system and

integrate it with other technologies to enhance utility and applicability in diverse
contexts.

5. User Training and Support: Provide comprehensive user training and support
resources to maximize user benefits and facilitate efficient system navigation.

BIBLIOGRAPHY

1. Ahmad, F., Wasi Abbas, S., Singh, J., & Kumar Mishra, N. (2015).

"Students Attendance Monitoring System Based on RFID and GSM

Network." Retrieved from https://www.jetir.org/papers/JETIR1504072.pdf

2. Kohli, S., Godwin, G. T., & Urolagin, S. (2020). "Sales prediction using linear

and KNN regression." In Advances in Machine Learning and

Computational Intelligence: Proceedings of ICMLCI 2019 (pp. 321-329).

Singapore: Springer Singapore.

3. Lalchandani, A., & Patel, S. (2021, September). "Smart IoT Based People

Counting System." In 2021 International Conference on Artificial

Intelligence and Machine Vision (AIMV) (pp. 1-6). IEEE.

4. Ikemura, S., & Fujiyoshi, H. (2011). "Real-time human detection using

relational depth similarity features." In Computer Vision–ACCV 2010:

18
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

10th Asian Conference on Computer Vision, Queenstown, New Zealand,

November 8-12, 2010, Revised Selected Papers, Part IV 10 (pp. 25-38).

Springer Berlin Heidelberg. Retrieved from

https://link.springer.com/chapter/10.1007/978-3-642-19282-1_3

5. Ma, H., Zeng, C., & Ling, C. X. (2012). "A reliable people counting system via

multiple cameras." ACM Transactions on Intelligent Systems and

Technology (TIST), 3(2), 1-22.

19
 SCIENCE, TECHNOLOGY and ENGINEERING PROGRAM

APPENDICES

20