VR - MALAZA - MILLAN - SAZON - Santiago - Chapter 1-3

Kayumanggi: An Immersive Virtual Reality Journey into the Lerion and Cariñosa Folk
Dances of High School Students in Marikina
Almilyn Mae C. Malaza, Arjet R. Millan, and John Joshua B. Sazon
College of Computer Studies, Technological Institute of the Philippines - Quezon City
IT 304: Capstone Project 1
Ms. Nila D. Santiago
July 2024
Kayumanggi: An Immersive VR Journey into the
Lerion and Cariñosa Folk Dances 2
Table of Contents
Table of Contents 2
List of Figures 3
List of Tables 4
Chapter 1 5
Introduction 5
Background of the Study 6
Project Objectives 8
Significance of the Study 9
Scope and Delimitation 10
Scope of the Study 10
Delimitation of the Study 11
Chapter 2 13
Theoretical Framework 13
Review of Related Literature 13
Conceptual Framework 28
System Architecture 29
Definition of Terms 30
Chapter 3 33
Research Methodology 33
Design Discussion 34
Design Trade-Offs 39
Project Development 48
Testing and Operating Features 50
Project Evaluation 50
Work Plan 52
Potential for Commercialization 52
Market Model 53
Measurable Benefits 54
Three-Year Product Roadmap 55
Go To Market Strategy 56
References 57
List of Figures
Figure 1 Example of Mature Dragon Fruit 14

Figure 2 Example of Immature Dragon Fruit 14
Figure 3 Different types of dragon fruit 18
Figure 4 Different types of diseases of dragon fruit 20
Figure 5 Design System 21
Figure 6 Image Acquisition and The Dataset Types 22
Figure 7 The Example Result of Pre-Processing phase 22
Figure 8 The Proposed of CNN Architecture 23
Figure 9 How the dataset is arranged 24
Figure 10 Dataset's augmented photos of dragon fruit 25
Figure 11 Techniques or algorithms to detect and classify fruit 25
Figure 12 Conceptual Framework 29
Figure 13 System Architecture 30
Figure 14 Design 1 36
Figure 17 Performance Ranking Scale 41
Figure 18 Render Time Ranking Scale of Design 2 and Design 1 42
Figure 19 Render Time Ranking Scale of Design 3 and Design 2 43
Figure 20 Memory Usage Ranking of Design 2 and Design 1 44
Figure 21 Memory Usage Ranking of Design 3 and Design 1 44
Figure 22 File Size Ranking of Design 2 and Design 1 45
Figure 23 File Size Ranking of Design 3 and Design 1 46
Figure 24 Agile Methodology 49
Figure 25 ISO 25010 52
Figure 26 Market Model 54
Figure 27 Go-to Market Strategy 57
Figure 28 The Business Model of DragonEye 57
List of Tables
Table 1 Summary of Constraints 39

Table 2 Design Trade-Off 1 41
Table 5 Criterion Performance Table 39
Table 6 Summary of Constraints/ Design Trade-Offs 46
Table 7 Work Breakdown Structure 52
Table 8 The Three-Year Product Road Map of DragonEye 55
Chapter 1
Introduction
Filipino folk dances, such as Lerion and Cariñosa, are essential parts of the
country's rich cultural heritage, connecting it to its past, traditions, and identity. They
represent customs passed down through the ages and the various cultural influences that
have shaped Filipino society over centuries, acting as significant repositories of cultural
heritage. These dances are highly regarded for their ability to capture the spirit of Filipino
customs, providing lively portrayals of principles, ideologies, and historical events that
profoundly impact Filipino communities and others around the world.
Cultural heritage encompasses more than physical monuments and objects; it
includes intangible elements like traditional dances, performing arts, social practices,
rituals, and festive events. According to No and Esguerra (2023), dance reflects one's
feelings, culture, history, and identity, and it is included in the Philippine education
curriculum to promote national identity and preserve cultural heritage through folk dances.
The decline in interest in traditional dances among students poses a threat to these
cultural legacies, as noted by Lobo (2022). Factors such as changing social norms, the
influence of contemporary dance styles, and the availability of traditional dance instruction
contribute to this decline. “Kayumanggi: An Immersive VR Journey into the Lerion and
Cariñosa Folk Dances” aims to address these challenges by using virtual reality technology
to create an immersive and authentic experience, enhancing understanding and appreciation
of these dances and ensuring their preservation for future generations.

Background of the Study
The traditional folk dances of the Philippines are a vital link to the history, customs,
and collective identity of the nation, and they are a rich store of cultural heritage. Lerion
and Cariñosa stand out among these dances because of their cultural and historical
significance. The aging of skilled dancers, the decline of traditional dance schools, and the
pervasive influence of modern entertainment forms pose significant challenges to the
transmission of these dances to younger generations. Innovative techniques are needed in
light of these problems in order to protect and promote these cultural treasures.
Pistola et al. (2019) demonstrated that cultural heritage encompasses far more than
just physical monuments and collections of objects. It also includes the traditional dances,
performing arts, social practices, rituals, festive events, and knowledge passed down from
our ancestors. These intangible elements are just as crucial to understanding and preserving
the unique identity of a community.
According to No and Esguerra (2023), dance is an art form that uses physical
movement to show one's feelings, culture, history, and other characteristics. It comes in
various styles, from the most traditional to the most contemporaneous. One is a folk dance
noted as sacred or secular, reflecting the nation's identity. Dance education can be found in
the curriculum of the Department of Education (DepEd) of the Philippines, under the K to
12 Program, designed to incorporate dancing into physical education subjects in the
classroom. Dance education, aside from health benefits, is to promote and help preserve the
national identity through its folk dances.

According to Lobo (2022), traditional dances are clearly losing appeal to students.
The rich cultural legacy that is ingrained in these dances is being threatened by this decline,
which brings with it a great challenge. Lobo highlights that more research is necessary to
determine the underlying causes of this decline in interest. Teachers and legislators can
create focused plans to reignite students' passion for dance by investigating elements like
shifting social norms, the impact of contemporary dance styles, and the availability of
traditional dance instruction. According to Lobo, implementing innovative teaching
methods and incorporating cultural education into the broader curriculum could help
students' appreciation for culture and the arts. Traditional dances can thrive and be
treasured by future generations if we cultivate a deeper appreciation and engagement.
Two individuals, attired in 19th-century Fil-Hispanic attire, exchange glances
across a room. They engage in a ritualistic circling before dancing around a handkerchief,
each holding opposite ends. This dance, known as the Cariñosa, epitomizes intense
romanticism: the man in relentless pursuit of the woman, while she responds with a blend
of encouragement and reserved demeanor. The Cariñosa is an integral component of the
Maria Clara Suite, a collection of Filipino folk dances that encapsulate themes of courtship
and cultural heritage (Diaz Rodriguez & Patrick, 2023).
According to Mariano (2024), Lerion, also known as Leron or Leron Sinta, is
celebrated as Marikina's folk dance, embodying themes of love and courtship deeply
intertwined with nature. This dance tradition reflects the community's affection for its
surroundings, emphasizing a cultural connection where love for nature is expressed
through rhythmic movements and symbolic gestures. Lerion serves as a cultural narrative,
illustrating how local traditions and values are preserved and celebrated through folk art
forms like dance, fostering a sense of identity and continuity within the Marikina
community.
The application provides a valuable opportunity for the preservation and
dissemination of folk dance traditions through the use of Virtual Reality (VR) technology.
By allowing users to immerse themselves in a simulated environment and observe, study,
and practice dance movements in a virtual space, VRTech offers a distinctive and creative
method for engaging with and preserving these cultural arts (Wang and Tseng 2023).
Preservation and promotion of traditional Philippine folk dances like Lerion and
Cariñosa are the goals of "Kayumanggi: An Immersive VR Journey into the Lerion and
Cariñosa Folk Dances" development. These dances are essential to the nation's cultural
legacy, but fewer traditional dance institutions and an aging dancer population make it
difficult to pass them on to future generations. Through the use of virtual reality
technology, this project aims to produce a compelling and genuine VR experience that
improves comprehension and appreciation of these dances, guaranteeing their accessibility
and preservation for future cultural enjoyment and education.
Project Objectives
The main objective of this study is to develop an immersive virtual reality (VR)
application titled “Kayumanggi: An Immersive VR Journey into the Lerion and Carinosa
Folk Dances,” aimed at preserving and promoting traditional Philippine folk dances
through cutting-edge VR technology.
Specifically, this project aims to:
● Developing a virtual reality application that simulates an immersive environment for
experiencing and learning traditional Philippine folk dances. This platform allows users
to engage deeply with Lerion and Cariñosa, providing a space for learning, practice, and
cultural appreciation in a virtual setting.
● Integrating 3D computer graphics for the simulation to provide a virtual environment and
use it as an interactive learning space for the students to help them understand Lerion and
Cariñosa folk dance.
● Evaluate the application using the ISO/IEC 25010 in terms of functionality suitability,
reliability, usability, performance efficiency, compatibility, maintainability, and
portability.
Significance of the Study
The project, entitled “Kayumanggi: An Immersive VR Journey into the Lerion and
Carinosa Folk Dances,” aims to develop a virtual reality application to preserve and
promote these traditional Philippine folk dances. The study holds significance for the
following stakeholders:
● Cultural Practitioners and Historians - This study provides an innovative tool to
preserve the intricate details of Lerion and Cariñosa, ensuring that these dances are
accurately documented and can be passed down to future generations.
● Educators and Students - The virtual reality application provides a unique and
engaging way for educators to teach and students to learn about Philippine folk
dances. It improves the educational experience by creating an immersive
environment in which users can interact with the dances in meaningful ways.
● Future Researchers/Developers -This study will provide future researchers with a
valuable resource to guide them in improving and expanding its capabilities, while also
gathering crucial information for future endeavors.

● Information and Communication Technology - The study exemplifies the successful
application of VR technology in cultural preservation, offering ICT professionals insights
into digital technology implementation in cultural contexts. It serves as an important
reference for advancing digital solutions that enhance cultural education and preservation
efforts, promoting the sustainable management of cultural heritage.
Scope and Delimitation
This study's scope outlines what is covered in the virtual reality under development,
while its delimitations specify what is not included.
Scope of the Study
The study focuses on the development of the Kayumanggi virtual reality application
with the following:
Main Menu
● Dance Simulation
○ This option allows users to enter the immersive virtual environments
where they can practice the precise movements of Lerion and Cariñosa
using advanced motion capture technology.
● Help
○ Provides instructions and guidelines on how to use and navigate through
the VR application.
➢ Introduction to Controls: Explains how to use VR controllers to
interact with the virtual environment, such as executing dance
steps and navigating menus.

➢ Objective Explanation: Clarifies the learning outcomes related to
mastering dance techniques, understanding cultural contexts, and
completing interactive tutorials.
➢ Progression Explanation: Guides users on advancing through the
dance simulations, improving dance skills, and accessing quizzes
and challenges.
● Settings:
○ Users can adjust the language preferences (English or Filipino) and audio
settings for a personalized experience.
● Exit:
○ An option to close the application.
Folk Dances
The modules featured in Kayumanggi: An Immersive VR Journey into the Lerion
and Cariñosa Folk Dances encompass the following:
● Learn Dances
○ Practice Lerion and Cariñosa using advanced motion capture in virtual
environments.
● Cultural Context
○ Explore the historical and cultural significance of Lerion and Cariñosa
through virtual exhibits.
● Interactive Learning
○ Step-by-step guides for hands-on learning of dance techniques in a virtual
space.
● Quizzes and Challenges

○ Includes quizzes and challenges for interactive and educational engagement.
● Performance Evaluation
○ Provides tools for users to evaluate their dance performances against the
captured motion data, allowing them to improve their skills and
understanding of the dances.
● About Module
○ Contains information about the development team, the project's mission,
goals, and contact details to build user trust and provide transparency.
Delimitation of the Study
● First-Person Perspective
○ Users experience the virtual environment strictly from a first-person
perspective, enhancing immersion in the dances of Lerion and Cariñosa.
● Focus on Dance Practices
○ The application only covers the steps, history, and cultural context of Lerion and
Cariñosa, excluding other dance forms and unrelated topics.
● Limited Movement
○ Interaction in the virtual space is limited to specific parameters, ensuring a
controlled environment for learning and practicing dance movements.
● Offline Access
○ Once installed, the application operates offline, enabling local access to
content, enhancing immersion in Lerion and Cariñosa dances.

Chapter 2
Theoretical Framework
This chapter focuses on the studies that contain facts and information about the research
problem. It also explains and logically connects the current study and earlier research.
Review of Related Literature
According to Deep, L. et al. (2022), fruits of the highest quality for consumers and
increased profitability for growers are harvested when they are at their ideal ripeness. For
the best fruit quality, fruits should be harvested at the right time and developmental stage.
A widespread practice in the international trade of numerous fruit crops is to harvest them
before they reach full maturity in order to obtain an extended storage life. For the sake of
trade, this frequently results in a compromise on the potential quality of the fruit in
question. Dragon fruit has maturation requirements, however they differ depending on the
region and cultivation circumstances.
A mature dragon fruit has an attractive appearance. It is hard, larger in size, has a
bright red or magenta hue (depending on the type), spiky skin, a sweet and somewhat
acidic flavor, well-developed seeds, and a sweet tropical smell when ripe. Scales or spikes
typically cover the skin, giving it an unusual and exotic appearance (Suitcase, 2023). These
visual and sensory characteristics are essential for consumers and play a crucial role in
quality grading. Identifying these features accurately is vital for ensuring that the fruit
meets market standards and consumer expectations.

Figure 1
Example of Mature Dragon Fruit.
Compared to its mature counterpart, immature dragon fruit is smaller, usually green
or light pink, firmer to the touch, milder and less sweet, contains underdeveloped seeds,
and sometimes tastes somewhat sour. Unlike fully mature dragon fruit, which is softer and
sweeter, this fruit is firmer. The precise traits may differ based on the type of dragon fruit
and its particular maturity stage (Horticult, 2022).
Figure 2
Example of Immature Dragon Fruit.

Additionally, the outer skin of an immature dragon fruit is a vibrant green during
the first stages of fruit growth and gradually turns red as the fruit ripens. The dragon fruit
ripens between 30 and 35 days after growing. An additional benefit of the crop is that
harvesting time can be adjusted based on market needs. Fruit for the local market can be
collected three to four days following the skin's transition from green to reddish-pink,
though it can be harvested a day earlier for export or overseas markets (Nangare, D. et al.,
2020).
Moreover, the qualities and freshness of dragon fruits are directly related to skin
look, color, texture, flavor, size, and shape. The original and expanded datasets of the
Dragon Fruit Maturity Detection Dataset are divided into two groups: Mature Dragon Fruit
and Immature Dragon Fruit. Similar to this, the original and enhanced datasets for the
Dragon Fruit Quality Grading Dataset are split into two groups: Fresh Dragon Fruit and
Defect Dragon Fruit. Relevant pictures of dragon fruits can be found in each of these
categories (Horticult, 2023).
Controlled variance is applied to the training dataset to improve the model's
flexibility in real-world scenarios. The key techniques are random rotation (0-45 degrees),
horizontal/vertical shifting (up to 20% width/height), and shearing for different viewpoints.
Horizontal flipping is useful for teaching orientation invariance; random zooming
(80–120%) is useful for handling different scales. In a fill mode, the brightness (50–150%)
and contrast (70–130%) are adjusted to allow for variations in lighting without affecting
the original content of the image. Pre-processing increases the model's adaptability by
adding a random zoom function. These tactics have allowed models to recognize objects in
a wide range of real-world circumstances. Figure 3 shows the dataset's augmented photos
of dragon fruit (Khatun, 2024).
According to Dhakal, A. et al. (2021), any commodity must first be quickly
commercialized, but it is necessary to determine how farmers view it and what amount of
knowledge they possess. It is only after identifying the farmers' perspectives and the basic
problems with adoption that tactics can be modified to address their concerns and improve
the harvest. In addition to identifying obstacles to the crop's approval, the study examined
the farmers' degree of awareness regarding dragon fruit and their preparedness for
adoption.
Dragon fruit farming was also seen to have a number of benefits by farmers,
including ease of cultivation, a short flowering period, year-round harvesting, minimal
harvest failure risk, and a fair market price. Social media was the second most important
source of information, after other farmers. Just 23.3% of farmers thought they were
proficient in growing dragon fruits, which highlights the need for them to continue
expanding their knowledge and expertise. For the district and province as a whole to
increase the production and sustainability of dragon fruit cultivation, the local government
should regularly offer extension services and other forms of assistance (Saediman, H. et
al., 2021).
The rapid expansion of planting area and production is the primary cause of the
major challenges facing Taiwan's dragon fruit market. Weak exports have resulted from
variable fruit quality and expensive production costs. Fruit quality and product safety must
be improved in order to become more competitive. On the other hand, pollination by hand
is necessary for the red-fleshed kinds, and the fruit's easy shattering is one of its many
drawbacks. The fruit comes in a variety of fruit sizes. The goal of the dragon fruit industry
lately has been to introduce or breed a spineless species with a flavor profile, distinct flesh
color, appropriate size, and extended shelf life. In order to increase the share of
high-quality fruits and supply consistency, the industry managers also intend to establish
quality production groups, support organic farming and the Traceable Agricultural
Products (TAP) system, develop facilities that culture technology throughout the supply
chain, and have Technical Service Groups conduct on-site demonstration sessions (Lee, W.
et al., 2020).
Vietnam's export of dragon fruit still encounters challenges. Firstly, higher quality
standards, technical guidelines, and inspections are being applied in most major
international markets, notably China, which puts pressure on Vietnam's dragon fruit
industry to raise the quality of its products. The second is the explosive growth of dragon
fruit production in China, India, and other Asian, Central American, and South American
nations. Limitations in the current quality infrastructure (certification services, testing) and
control of farming methods may present further issues (Son, N. et al., 2020).
Due to its low production and export volume, dragon fruit is not considered a major
fruit variety. Dragon fruit is exclusively produced in Malaysia at a rate of 0.6% of total
fruit production, with Singapore being the primary destination for exports (not exceeding
2% of total exports). Diseases and pests are a couple of challenges faced in growing
dragon fruit. (Sarip, 2020). According to Kumar et al. (2023), manual detection of dragon
fruit stem diseases for large scale farming is difficult and depends on the expertise of the
farmer on diseases.
Furthermore, four types of dragon fruits are currently contributed to the global
market: (i) red skin, white flesh (Hylocereus undatus), primarily from Vietnam and
Thailand; (ii) red skin, red flesh (Hylocereus polyrhizus), primarily from Malaysia and
Israel; (iii) red skin, purple flesh (Hylocereus costaricensis), primarily from Guatemala,
Nicaragua, Ecuador, and Israel; and (iv) yellow skin, white flesh (Ecuador and Colombian
Hylocereus (Selenicerus) megalanthus). In the global market, the shares of red skin with
white flesh, red skin with purple flesh, red skin with red flesh, and yellow skin with white
flesh are roughly 94, 4.0, 1.5, and 0.5%, respectively. Based on estimates, the global output
of dragon fruit is estimated to reach above 2.1 million tons (Mt) in an area of 1.12 lakh
hectares in 2017–18. Taiwan, Malaysia, Thailand, Vietnam, China, and Indonesia. The top
producers are the USA, India, Cambodia, and the Philippines (Wakchaure, GC. et al.,
2021). See figure 3.
Figure 3
Different types of dragon fruit.

However, in spite of its significant production loss, fungus disease is one of the
main issues affecting dragon fruit production's yield, quality, and profit today. The study
revealed fungal infections that affect dragon fruit (Hylocereus undatus) plants in the
southern Philippines. Samples were gathered on a few farms in the Philippines' Saranggani
Province. The dragon fruit plants' diseased sections were gathered and taken to the lab so
that the fungus that triggers the disease could be grown there. The detached stem method
was utilized to determine the infectious capacity of grown fungus after they were
pure-cultured. The first important step in creating solutions to problems relating to fungal
diseases is identifying fungal pathogens on economically important crops. This will assist
farmers to increase their yield. (Emejas Jr., D. P. et al., 2023).
According to Balendres and Bengoa (2019), there are 21 fungal species connected
to diseases affecting dragon fruit. The most serious of these that affect the fruit and stem is
anthracnose. After harvest, anthracnose can happen in the field. Anthracnose is identified
by its characteristic reddish-brown irregular or circular spots that either emerge on the fruit
or the stem. These spots eventually merge, grow, and transform into dark brown sunken
lesions. Conidial masses form in these sunken lesions, and chlorotic shadows encircle the
lesions. While anthracnose cases have been documented in a number of producing nations,
there haven't been many reports from Vietnam, Indonesia, or Sri Lanka—three of the
leading dragon fruit producers. See figure 4.

Figure 4
Different types of diseases of dragon fruit.
Figure 4. Anthracnose of dragon fruit. (A) Stem anthracnose. (B) Fruit anthracnose.
Besides, stem rot (stem anthracnose) disease caused by Fusarium spp., has been
documented in Indonesia and Malaysia. Brown circular sunken sores with white mycelia
and orange sporodochia are the etiological symptoms of infected Hylocereus polyrhizus
(Hylocereus monacanthus). Also, the most common postharvest fruit rot of dragon fruit is
anthracnose, which is brought on by the Colletotrichum species. Fusarium spp., and
Bipolaris cactivora are also present. have been known to cause dragon fruit to turn brown.
Wounds and other predisposing conditions are common causes of infection. (Zakaria,
2022).
The Convolutional Neural Network (CNN) is used in model building since it is an
effective way to classify images using a lot of datasets and class labels. The study's idea is
that disease data on dragon variant imaging can be more accurately classified using the
suggested method. Convolutional Neural Networks with data-adapted architectures were
employed in the suggested study approach to classify a few dragon fruit disorders. It
illustrates the many stages of the overall approach, which include image acquisition and
data input, pre-processing, Convolutional Neural Network (CNN) method training and
testing, and evaluation. (Hakim, L. et al., 2023). See figure 5.
Figure 5
Design System
Data on the type of diseases on dragon fruit plant stems is now being gathered as
much as feasible. The jpg/jpeg format of the image data, distinct camera angles, the same
shooting distance, and clear, unblurry photographs are required for this data collection
procedure. Out of all the data, 10% is used for test data, 10% is used for data validation,
and 80% is used for training data. The five desirable disease forms in this study are mosaic
(E), black rot (C), red spot (A), stem rot (B), and scabies (D). (Hakim, L. et al., 2023). See
figure 6.
Figure 6
Image Acquisition and The Dataset Types
In addition, the pre-processing phase, also called data augmentation, is the stage of
picture preparation that comes before additional processing. In data augmentation,
different techniques are applied to ensure that the data class labels stay stable while
introducing random disruptions and creating a "new" training sample from the original.
The next step can be accelerated by modifying the picture pixels (target size) in the Keras
picture Data Generator at this stage. Next, carry out a zoom range, shear range, and
horizontal flip. (Hakim, L. et al., 2023). See figure 7.
Figure 7
The Example Result of Pre-Processing phase

After images successfully pass the pre-processing step, they are classified using the
CNN technique, which divides input images into discrete groups depending on their output
values. The training and testing data are processed using the CNN technique. During the
data training phase, the CNN technique produces the most accurate model for classifying
the various diseases on dragon fruit stems. This model will serve as the basis for the data
testing process, together with extra test data. Feature learning and classification are the
two main processes in the CNN approach. With the use of a technique called feature
learning, a system may be programmed to automatically translate a picture's representation
into features—numbers that stand in for the image. (Hakim, L. et al., 2023). See figure 8.
Figure 8
The Proposed of CNN Architecture
However, using MobileNetV2 as a feature extraction block to reduce model
parameters, focus loss rather than cross entropy loss, and an improved feature fusion
structure. According to Zha, et al. (2021) proposed the YOLOv4_MF model for creature
detection. In the end, the model's mAP was 4.24% higher than YOLOv4, and its volume
was reduced to 1/6 of YOLOv4.
Regarding real-time fruit maturity and localization, this is a study that suggests an
efficient, enhanced YOLOv5 (You Only Look Once) based algorithm. Initially, this
technique rebuild the YOLOv5 backbone network using the MobileNetV3 back neck
module, replacing the original focus layer with a down sampling convolutional layer. It
then employs a genetic algorithm for hyperparameter tuning to increase detection accuracy
and performs out channel trimming for the neck layer to further minimize the model size.
(Zeng, T. et al., 2023).
For deep learning models, particularly in visual object recognition, data
augmentation is essential. Specifically, it augments the training dataset by generating new
images from the ones that already exist, improving model generalization, and decreasing
overfitting. It is an effective tool for strengthening deep learning models (Khatun, T. et al.,
2024). It also presented a deep learning model that aims to achieve state-of-the-art results
by efficiently training the dataset. A comprehensive assessment of this deep learning
model's performance on a dataset is necessary for its validation. An interconnected layer of
nodes, representing individual processing units, makes up a deep-learning model. The
nodes in the input layer receive data, and the nodes in the output layer produce the final
result. These input and output layers are separated by hidden layers, which contain the
neural network's main processing power (Mesopotamian, 2022).
Moreover, in the analysis of visual data, deep learning models have achieved
significant progress in tasks including object detection, natural language processing, and
picture or video classification (Islam, M.A. et al., 2021).

Using a dataset of images of dragon fruits, the suggested approach uses various
machine learning algorithms, including Artificial Neural Network (ANN), Support Vector
Machine (SVM), and Convolutional Neural Network (CNN), to detect dragon fruit
accuracy. It also sorts the infected, high, medium, and low-quality dragon fruits using
Raspberry Pi functionality (Patil, P. et al., 2021). See figure 9, 10 and 11.
Figure 9
It shows how the dataset is arranged.
Figure 10
It shows the dataset's augmented photos of dragon fruit.

Figure 11
Examples of techniques or algorithms to detect and classify fruit quality using various features
of fruits and vegetables.
This means that there are currently numerous methods or algorithms for identifying and
categorizing fruit quality based on distinct fruit and vegetable characteristics (Patil, P. et al.,
2021).
Synthesis
Dragon fruit quality and optimal harvest time are crucial for both consumer
satisfaction and grower profitability. The fruit's visual and sensory attributes, such as size,
color, sweetness, and aroma, significantly influence its market acceptance and grading
standards (Deep et al., 2022; Suitcase, 2023). Differentiating between mature and
immature dragon fruits involves assessing distinct characteristics in size, color, taste, and
seed development (Horticult, 2022). Harvesting practices vary, with considerations for
local versus export markets impacting the timing of harvest (Nangare et al., 2020).
Ensuring the quality and freshness of dragon fruits remains critical, with advanced
datasets and machine learning techniques playing a pivotal role in maturity detection and
quality grading (Horticult, 2023; Khatun, 2024). Despite benefits such as year-round
cultivation and reduced harvest risks, challenges persist in dragon fruit farming, including
variable quality, high production costs, and competitive market conditions (Dhakal et al.,
2021; Lee et al., 2020). Technological innovations like the DragonEye mobile application,
utilizing CNN models for automated maturity detection and quality grading, aim to address
these challenges by enhancing efficiency and accuracy in assessing fruit quality, thereby
bolstering market competitiveness and meeting evolving consumer expectations.
Furthermore, collaborative efforts between researchers, growers, and technology
developers are essential for advancing these innovations and ensuring sustainable practices
in the dragon fruit industry (Gupta et al., 2023). By integrating these advancements,
stakeholders aim to optimize crop management practices, minimize environmental impact,
and enhance overall profitability in dragon fruit cultivation.

Conceptual Framework
Figure 12
Conceptual Framework of DragonEye
Figure 12 illustrates the conceptual framework guiding the development of
"DragonEye: A Mobile Application for Maturity Detection, Disease Detection and Quality
Grading of Dragon Fruit Using Convolutional Neural Network (CNN) MobileNetV3
Model." This framework is structured into Input, Process, Output, and Evaluation. Input
encompasses the knowledge requirements for Python and Dart programming languages,
understanding of dragon fruit maturity detection, familiarity with CNN’s MobileNetV3
model, and required software and hardware for mobile application development. In the
Process, DragonEye will adopt Agile Methodology, proceeding through 6 phases:
Planning, Design, Development, Test, Deployment, and Review. The Output will be the
DragonEye application, which will undergo evaluation based on ISO 25010 criteria,
assessing functional suitability, reliability, performance efficiency, usability, security,
compatibility, maintainability, portability.
System Architecture
Figure 13
System Architecture of DragonEye
Figure 13: The DragonEye mobile application system architecture is composed of
three primary tiers: Presentation, Logic, and Data. In the Presentation Tier, users interact
with the app on their mobile devices. This interface enables farmers to capture images of
dragon fruits for maturity detection, disease detection and quality grading. The Logic Tier
handles the processing of these images using advanced machine learning models. This
involves employing object detection and data analysis techniques to assess the maturity and
quality of the dragon fruits. Lastly, the Data Tier is responsible for storing and retrieving
information related to dragon fruit characteristics. This data assists the application in
providing accurate assessments by comparing the captured images with the stored datasets.
DragonEye leverages mobile technology and machine learning to offer a user-friendly and
effective tool for optimizing crop management and ensuring high-quality produce.
Definition of Terms
Agile Methodology - A flexible project management approach emphasizing iterative
development and adaptability to changing requirements, commonly used in software
development.
Anthracnose - A group of fungi-related diseases, mostly from the genus Colletotrichum,
that are brought on by various fungal species. Many different types of plants are affected by
anthracnose, which causes black, sunken lesions on leaves, stems, flowers, and fruits.
These lesions have the potential to enlarge and cause fruit rot, leaf drop, and general plant
damage.
Artificial Neural Network (ANN) - A computational model inspired by biological neural
networks, utilized in machine learning for tasks such as pattern recognition and regression
analysis.
CNN (Convolutional Neural Network) - A deep neural network designed to process
grid-like data like images and videos, enabling tasks such as image classification and object
detection.
Conceptual Framework - A structured plan guiding the development of systems or
applications, outlining inputs, processes, outputs, and evaluations to achieve specific goals.
Data Augmentation - Techniques to enhance datasets by generating new variations of
existing data, improving the training of deep learning models for better generalization and
robustness.
Dragon Fruit - A tropical fruit with varieties including red and yellow-skinned types with
white, red, purple, or yellow flesh, cultivated globally.
Fruit Rot (Fruit Anthracnose) - A kind of anthracnose that fungus-like Colletotrichum
produces to damage plant fruits. The disease causes black, sinking lesions or patches on the
fruit's surface. These lesions might enlarge and lead to fruit rot, which will lower the fruit's
quality and decrease its market value.
Fungal Disease - A kind of plant disease brought on by pathogenic fungus that can infect a
plant's leaves, stems, roots, and fruits, among other sections of the plant. These diseases
frequently manifest as spots, blights, rots, and dying, and they have the potential to harm
plants as well as lower their yield and quality.
ISO 25010 - Standards for evaluating software quality attributes such as functionality,
reliability, performance, usability, security, compatibility, maintainability, and portability.
Machine Learning - AI techniques enable computers to learn from data and make
predictions or decisions without explicit programming.
MobileNetV3 - A variant of MobileNet optimized for mobile and embedded vision tasks,
known for its efficiency in image recognition and object detection.
Object Detection - Computer vision technique identifying and locating objects in images
or videos, crucial for applications like autonomous systems and surveillance.
Quality Grading - Assessment of agricultural produce based on factors like appearance,
texture, taste, and nutrition, crucial for competitiveness in the market.
Stem Rot (Stem Anthracnose) - A type of anthracnose that attacks plant stems and is
frequently brought on by fungi belonging to the Colletotrichum genus. It appears as dark,

sunken lesions on the stems that, in extreme cases, can wrap and choke the stem, resulting
in drooping and dieback.
System Architecture - Structure of a software system or application defining components,
interactions, and dependencies to achieve specific functionalities.
Training Dataset - Collection of data samples used to train machine learning models,
essential for model development, optimization, and validation.
YOLOv8 (You Only Look Once version 8) - It is an advanced real-time object detection
algorithm that utilizes deep learning and convolutional neural networks to efficiently and
accurately detect and classify objects in images, known for its improved performance in
handling object sizes and precise localization.

Chapter 3
Research Methodology
This chapter discusses the primary approach, which focuses on using a
developmental research process for project development. It consists of stages such as
project design, development, testing, and evaluation, to establish a functional system that
meets requirements and improves productivity and user experience. This study will follow
a primary research methodology and approach aimed at expanding and enhancing existing
knowledge, particularly in the field of computer vision. The goal is to explore its full
potential for innovation within current technological capabilities.
Project Design
DragonEye aims to develop a smartphone application that utilizes a camera to
assess the maturity, disease, and quality of dragon fruits. This application gives farmers the
ability to quickly identify the ripeness and quality grade of their produce. It enables them to
improve crop health, reduce losses from suboptimal harvests, and lower operational costs
related to quality control. This includes addressing complications in crop management and
improving overall efficiency in dragon fruit farming. In addition to maturity detection and
quality grading, the app offers users the opportunity to detect fungal diseases such as Stem
Anthracnose, Fruit Anthracnose, Stem Rot, and Fruit Rot. The camera function of mobile
devices allows users to capture images of the fruits and assess their condition. If an issue is
identified, the app will provide farmers with the necessary information for effective disease
management by offering details on the fruit's condition, symptoms of any problems, and
suggested solutions.
Design Discussion
DragonEye aims to develop a smartphone application that utilizes a camera to
assess the maturity and quality of dragon fruits. This application empowers farmers to
efficiently identify the ripeness and quality grade of their produce, facilitating improved
crop health, reduced losses from suboptimal harvests, and lowered operational costs
associated with quality control. It addresses complexities in crop management and
enhances overall efficiency in dragon fruit farming. In addition to maturity detection and
quality grading, the app provides farmers with opportunities to detect fungal diseases such
as Stem Anthracnose, Fruit Anthracnose, Stem Rot, and Fruit Rot. Leveraging mobile
device cameras, farmers can capture images of the fruits and assess their condition. Should
any fungal diseases be detected, the app offers detailed information on the symptoms and
recommended measures for disease management. These features aim to equip growers with
the tools needed to proactively manage fungal diseases, optimize their farming practices,
and maximize yield and quality. Furthermore, DragonEye intends to integrate intuitive
interfaces and real-time data processing capabilities into the application. This design will
allow farmers to make informed decisions promptly, enhancing their ability to respond to
changing crop conditions and market demands. By promoting knowledge-sharing and
leveraging innovative technology tailored to the specific needs of dragon fruit cultivation,
DragonEye seeks to empower farmers and advance agricultural practices in the industry.
Figure 14
Design 1
For Design 1, the background features a linear gradient blending teal and dark blue,
lending a modern and appealing look. The top of the design showcases the DragonEye logo,
followed by a welcoming message that invites users to explore the app. Below the title, five
large, vertically stacked buttons with ample spacing guide the user through the app's key
functionalities: "Maturity Detection" helps assess dragon fruit maturity through captured or
uploaded pictures, "Quality Detection" evaluates the quality of the fruit, "Classify Variety"
identifies different varieties, "Learning Module" provides educational content related to dragon
fruit, and "About Us" offers brief information about the app and its creators. Each button is
adorned with relevant dragon fruit imagery and is easily accessible, ensuring a user-friendly
experience that emphasizes the app's cohesive theme and functionality.
Figure 15
Design 2
Design 2 showcases a user-friendly layout that ensures seamless navigation. At the top of
the interface, the DragonEye logo is prominently displayed, featuring a vibrant design that
encapsulates the essence of the application. Directly below the logo, the welcoming message
"Welcome to DragonEye!" invites users to explore the app. The main menu consists of five
square buttons, each depicting a specific functionality with both an icon and a label. The
"Maturity Detection" button, featuring an image of dragon fruits in a field, allows users to
determine the maturity level of the fruit. The "Quality Detection" button, adorned with an image
of sliced dragon fruit, helps users assess the quality of the fruit. The "Classify Variety" button,
with an image of different dragon fruit varieties, facilitates the identification of various types.
The "Learning Module" button, illustrated with dragon fruits on a plant, provides educational
content about dragon fruit cultivation and care. Lastly, the "About Us" button, depicted with an
icon of people, offers information about the application and its creators. This design ensures that
users can effortlessly access each feature of the DragonEye app, fostering a cohesive and
engaging experience.
Figure 16
Design 3
Design 3 presents a clean and inviting layout with a white background that emphasizes
simplicity and clarity. At the top, the DragonEye logo is prominently displayed, capturing the
app's focus on dragon fruit analysis. Below the logo, the welcoming message "Welcome to
DragonEye!" invites users to explore the app. The main menu features five large, circular
buttons, centrally aligned and spaced out for easy navigation. Each button includes relevant
imagery and text indicating its functionality: "Maturity Detection" with an image of dragon fruits
in a field allows users to assess ripeness; "Quality Detection," showcasing sliced dragon fruits,
helps evaluate quality; "Classify Variety," displaying various types, aids in identifying different
varieties; "Learning Module," illustrated with dragon fruits on a plant, provides educational
content on cultivation and care; and "About Us," depicted with an icon of people, offers
information about the app and its creators. This intuitive design ensures users can effortlessly
access and utilize the DragonEye app's features, making navigation straightforward and
engaging.
Design Trade-Offs
Table 1
Summary of Constraints
Design Constraints
Models
Render Time Memory Usage File Size
Design 1 71 ms 415 mb 145 kb

Equation 1
Percentage Difference
(𝐻𝑖𝑔ℎ𝑒𝑟 𝑉𝑎𝑙𝑢𝑒 − 𝐿𝑜𝑤𝑒𝑟 𝑉𝑎𝑙𝑢𝑒)

% 𝑑𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑐𝑒 𝑓𝑜𝑟 𝑀𝐼𝑁 = 𝐻𝑖𝑔ℎ𝑒𝑟 𝑉𝑎𝑙𝑢𝑒
Equation 1 computes the percentage
difference for each design by taking the difference between the higher and lower values and
dividing it by the higher value, according to the specified criteria.
Equation 2
Subordinate Rank
𝑆𝑢𝑏𝑜𝑟𝑑𝑖𝑛𝑎𝑡𝑒 𝑅𝑎𝑛𝑘 = 𝐺𝑜𝑣𝑒𝑟𝑛𝑚𝑒𝑛𝑡 𝑅𝑎𝑛𝑘 − [(%𝑑𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑐𝑒) * 10]
Equation 2 explains how the subordinate ranking of the three designs is determined by
subtracting the percentage difference from the Government Rank, multiplied by ten. The rank is
adjusted based on the magnitude of the percentage difference, with larger differences resulting in
greater deductions.
Figure 17
Performance Ranking Scale
This figure illustrates the performance ranking scale that will be employed to compare
the two designs based on their subordinate ranks (refer to Figure 17 for details).
Trade-Off 1: Render Time
Table 2 compares the render times of the three designs. Design 1 had the fastest loading
time, identified as the controlling benchmark with a ranking score of 10.

Table 2
Design Trade-Offs 1
Design Render Time
Design 1 71 ms
Design 2 76 ms
Design 3 70 ms
Computation of Ranking for Trade-Offs between Design 3 and 1
(71−70)
%𝑑𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑐𝑒 𝑓𝑜𝑟 𝑀𝐼𝑁 = 71
%𝑑𝑖𝑓𝑓𝑒𝑟𝑒𝑛𝑐𝑒 = 0. 14
Subordinate Rate = 10 -[(0.14) * 10]

Subordinate Rank = 8.6
Figure 18
Render Time Ranking Scale of Design 2 and Design 1
Figure 7 illustrates the secondary ranking of the render time for design 1, which is valued at 8.6,
compared to design 3's render time of 10 (refer to Figure 18).
Computation of Ranking for Trade-off 1 between Design 3 and Design 2:
(76 − 70)
Subordinate Rank = 10 -[(0.07) * 10]

Figure 19
Render Time Ranking Scale of Design 3 and Design 2
Figure 8 depicts the secondary ranking of the render time for design 2, which is valued at 9.4, in
comparison to design 3's render time of 10 (refer to Figure 19).
Trade-Off 2: Memory Usage
Table 3 presents a comparison of the memory usage among the three designs. This metric
indicates the amount of memory each design consumes during execution. Design 2, which
utilizes the least memory, was anticipated to be the leading rank and was assigned a value of 10.
Table 3
Design Trade-Off 2
Design Memory Usage
Design 1 415 mb
Design 2 437 mb
Design 3 420 mb
(437− 415)

Figure 20
Memory Usage Ranking of Design 2 and Design 1
Figure 9 shows a graphical representation where the memory usage of design 2, valued at
9.5, is plotted against the memory usage of design 1, which has a value of 10 (refer to Figure 20).
(420 − 415)

Figure 21
Memory Usage Ranking of Design 3 and Design 1
Figure 10 illustrates a graph where the memory usage of Design 3, valued at 9.9, is
plotted against the memory usage of Design 1, which has a value of 10. Design 1 exhibits the
lowest memory usage among the three designs (refer to Figure 21).
Trade-Off 3: File Size
Table 4 presents a comparison of the file sizes among the three designs. This metric
indicates the storage space consumed by each design in kilobytes. Design 1, which has the
smallest file size, was anticipated to be the leading rank and was assigned a value of 10.
Table 4
Design Trade-Off 3
Design File Size
Design 1 145 kb
Design 2 152 kb
Design 3 167 kb
(152 − 145)

Figure 22
File Size Ranking of Design 2 and Design 1
Figure 11 illustrates a graph where the file size of Design 2, valued at 9.5, is plotted
against the file size of Design 1, which has a value of 10 (refer to Figure 22).
(167 − 145)

Figure 23
File Size Ranking of Design 3 and Design 1
Figure 12 shows a graph where the file size of Design 3, valued at 8.7, is plotted against
the file size of Design 1, which has a value of 10 (refer to Figure 23).
Table 5
Criterion Performance Table
Decision Criterion’s Ability to satisfy the criterion (-10 to 10)
Criteria Performance
Design 1 Design 2 Design 3
Render Time 10 8.6 9.3 10
Memory Usage 8 10 9.5 9.9
File Size 9 10 9.5 8.7
Overall Rank 256 254.5 257.5

Influence of Design Trade-Offs in Final Design
Design trade-offs were employed to evaluate the efficiency and effectiveness of the
application's capabilities. Criteria were prioritized according to their significance. Render time,
memory usage, and file size were used as metrics to assess and compare the performance of the
designs generated. Consequently, Design 1 was chosen as the ultimate design.
Table 6
Summary of Constraints/Trade-Offs
Algorithm Constraints
Models
Strength Weaknesses Accuracy
YOLOV8 Fast detection speed Extending the 81.4%
and lower training period and
computational cost, demanding more
with high accuracy for iterations in order to
complex scenes. achieve the best
results.
SSD Faster detection speed Lower accuracy 77.3%
and lower compared to more
computational cost, complex models,
suitable for real-time struggles with small
applications. object detection.
Faster R-CNN High accuracy, Slower detection 80.2%

especially for complex speed and higher
and cluttered scenes, computational cost,
with strong requiring powerful
performance on hardware.
various tasks.
The table covers the three algorithm models: object detection models: YOLOv8, SSD,
and Faster R-CNN, focusing on their strengths, weaknesses, and accuracy to determine their
suitability for various applications. YOLOv8 stands out for its fast detection speed and lower
computational cost while maintaining high accuracy in complex scenes, making it ideal for
real-time applications requiring both rapid processing and precision. However, it necessitates an
extended training period and more iterations to achieve optimal results, which can be a drawback
in resource-constrained environments. YOLOv8 achieves an accuracy of 81.4%. Studies by
Redmon and Farhadi (2018) and Jocher et al. (2022) highlight YOLOv8's advancements in
architecture, improving both mean Average Precision (mAP) and inference time compared to
earlier YOLO models. SSD (Single Shot Multibox Detector) is recognized for its swift detection
speed and low computational cost, making it suitable for real-time applications and embedded
systems with limited computational resources. Despite these advantages, SSD's accuracy is
generally lower, particularly for small object detection, with an accuracy of 77.3%. Liu et al.
(2016) and Wang et al. (2020) discuss SSD's efficiency and its limitations in detecting objects at
various scales, impacting its overall precision. Faster R-CNN (Faster Region-based
Convolutional Neural Network) is noted for its high accuracy, especially in complex and
cluttered scenes, making it robust for tasks requiring high precision. The trade-off for this
accuracy is slower detection speed and higher computational cost, necessitating powerful
hardware. Faster R-CNN achieves an accuracy of 80.2%. Ren et al. (2015) and Huang et al.
(2017) emphasize Faster R-CNN's superior accuracy and its integration of a Region Proposal
Network (RPN) with Fast R-CNN, significantly improving detection performance.
The selection of the appropriate object detection model depends on specific application
requirements. YOLOv8 is optimal for applications needing both speed and high accuracy. SSD is
suitable for real-time applications with limited computational resources, and Faster R-CNN is
ideal for tasks requiring high precision in complex environments. Future research should aim to
develop models that combine the strengths of these approaches while mitigating their
weaknesses.
Project Development
Figure 24
Agile Methodology
Figure 24 outlines the flow of the Agile methodology process. The six (6) phases of agile are
plan, design, develop, test, deploy, and review. (see Figure 14)
Planning Requirements Phase: This stage, essential to the project development process,
is the planning phase. It involves gathering data, including pertinent information,
requirements, and limitations, to guide the project's direction. Emphasis is placed on
clarifying the project's goals and intended outcomes through defining the problem
statement and objectives to ensure a thorough understanding
User Design Phase: Based on the data collected and the review of relevant literature,
emphasis is placed on determining the system architecture. Initial models and prototypes
are developed based on this input. This phase involves iterative revisions until the system
architecture and project design align with the project's requirements.
Develop:.Once the system architecture and planning are done, the construction phase is
where the actual development, testing, and integration of the mobile application takes
place. Alpha testing and beta testing are the two testing steps that are needed in the
construction phase. The project team and the target users will assess and test the
prototypes. Further improvements and updates will be applied to the prototypes based on
the evaluation gathered by the project team and target users' assessment.
Deploy: This phase signifies the final stage of the project development, known as the
implementation phase, where the completed project is launched. Targeting farmers and
consumers as the primary users, testing and evaluation by this demographic will occur.
Ongoing testing by both developers and clients allows for final adjustments to be made to
the mobile application.

Testing and Operating Features
The DragonEye application will undergo thorough testing using Android
smartphones. Initially, developers will assess its functionalities to ensure they meet the
specified user requirements for evaluating dragon fruit maturity, disease and quality. Then,
farmers, who are the primary target users, will test the application's ability to accurately
determine the ripeness, fungal disease and quality grade of dragon fruits using their
smartphones. They will provide feedback to evaluate whether the application effectively
meets their needs in real-world farming conditions. Upon successful completion of these
evaluations, the application will be prepared for launch and operational use by farmers,
facilitating enhanced crop management and quality control.
Project Evaluation
To comply with the ISO/IEC 25010 standard, conducting user evaluations is
essential. Feedback and evaluations received during future project assessments will guide
enhancements and improvements to the mobile application.

Figure 25
ISO 25010
The evaluation of DragonEye will adhere to ISO 25010, a comprehensive standard
for assessing software quality across eight critical dimensions. This framework ensures that
DragonEye meets high standards in Functional Suitability, Reliability, Performance
Efficiency, Usability, Security, Compatibility, Maintainability, and Portability. By following
ISO 25010, DragonEye aims to ensure excellence in functionality, reliability, efficiency,
usability, security, compatibility, maintainability, and portability. This adherence enhances
the application's quality, security measures, adaptability, and maintenance across different
platforms and environments, bolstering its resilience and effectiveness in supporting
farmers' needs and advancing agricultural technology.

Work Plan
Table 7
Work Breakdown Structure
Potential for Commercialization
The DragonEye application is primarily designed to assist farmers in assessing the
maturity and quality of dragon fruits. Developed for mobile Android devices, this application can
be easily installed and accessed on smartphones. DragonEye serves as a valuable tool that
enables users to determine the ripeness and quality grade of dragon fruits, thereby enhancing
crop management and ensuring high-quality produce. After the development phase, the project
aims to make the application widely accessible, with the potential for commercialization focused
on benefiting the agricultural community rather than maximizing profit. The study's objective is
to support farmers by providing an essential tool that improves efficiency and productivity in
dragon fruit farming.

Market Model
Figure 26
Market Model
Figure 6 illustrates the Market Model detailing the global agricultural workforce,
focusing on the Philippines and CALABARZON region, outlining the Total Available
Market (TAM), Served Available Market (SAM), and Share of Market (SOM). According
to Mehrabi, 'Currently, around 600 million farms support 8 billion people globally, with
projections suggesting fewer farmers will feed a growing population in the future,
emphasizing the need for robust support systems' (Mueksch, 2023). According to Balita
(2024), The Philippines recorded 10.66 million agricultural workers in 2021, including 2.4
million farmers with an average age of 56, as reported by the Daily Tribune (2020).
CALABARZON hosts 140,000 agricultural workers across 435 establishments, defining
DragonEye's SAM and initial SOM targets according to Philippine Statistics Authority
(2023). This strategic approach aims to optimize agricultural productivity and management
within these specific regions, aligning with evolving needs and opportunities in Philippine
agriculture.
Measurable Benefits
The development of DragonEye, a smartphone application created by agricultural
technology researchers, offers significant measurable benefits. DragonEye uses advanced
image processing and smartphone camera technology to improve the detection of maturity
stages, disease and quality grading in dragon fruit. The main users of this app are farmers
and agricultural professionals. DragonEye makes it easier to assess dragon fruit quality,
enabling timely harvesting and effective quality control. This helps prevent financial losses
due to spoilage or poor quality, enhancing the economic success of dragon fruit farming.
The application provides farmers with a convenient way to monitor their crops
using their smartphones. DragonEye can also be used by agricultural professionals during
routine inspections to provide thorough evaluations without the need for expensive
equipment.
In conclusion, DragonEye represents a significant advancement in agricultural
technology, offering substantial improvements in fruit quality evaluation and management.
By enabling more accurate and timely assessments, DragonEye helps farmers maximize
yield and revenue while meeting consumer demand for high-quality produce. This
technology supports sustainable farming practices, reduces waste, and democratizes
advanced agricultural tools, benefiting both small-scale farmers and large enterprises.
Three-Year Product Roadmap
Table 8
The Three-Year Product Roadmap of DragonEye
Year 1 Year 2 Year 3
Price Free Free available on Free

Google Play and
AppStore
Performance It can be used on It can be used on It can be used on all

Android devices Apple devices and devices, future
launched on the App windows updates, and
store, mobile devices.
It can be downloaded
on the website
created for future
development
Features The DragonEye Improves the The app can cover

application can accuracy and different fruits
detect the maturity efficiency of maturity and quality
stage and quality classifying dragon grading.
grade of dragon fruit maturity and
fruits. quality using
advanced object
detection technology.
Go To Market Strategy
Figure 27
Go-to-Market Strategy
Offerings Customer Channels
Free download of the mobile Farmers and Consumers Google Play Store
application
The researchers chose to release the DragonEye mobile application on the
Android Marketplace, also known as the Google Play Store, a platform where Android
phone users can download the application.
Business Model
Figure 28
The Business Model of DragonEye

The business model for DragonEye illustrates its viability and impact potential. The
application aims to revolutionize dragon fruit quality evaluation by leveraging advanced
image processing and smartphone technology for automated maturity detection, disease
detection and quality grading. DragonEye will be launched on the Google Play Store,
offering free access to farmers and agricultural professionals. It utilizes sophisticated image
processing algorithms to assess the quality of dragon fruit, providing real-time feedback
through smartphone functionalities. DragonEye ensures Functional Suitability, Reliability,
Performance Efficiency, Usability, Security, Compatibility, Maintainability, Portability
adhering to ISO 25010 standards.
Customer relations will be fostered through feedback mechanisms, social media
engagement, and collaborations with agricultural organizations. Dissemination channels
include a dedicated website, social media platforms, and word-of-mouth recommendations.
The cost structure includes development, maintenance, and promotional expenses, aimed at
continuous improvement and user accessibility. DragonEye remains committed to its
mission of enhancing agricultural productivity and product quality, offering essential
evaluation services without cost to users.
References
VR FOLK DANCE
No, J. G., & Esguerra, V. B. (2023). Dance workshop for the preservice elementary teachers:
Reinforcing perceived teaching skills in the arts. Performance and Evaluation
Management Journal, 8(488-494). https://doi.org/10.5281/zenodo.7876658
Lobo, J. (2022). Do our folk dances still thrive? Personal experience and interest of students
towards Philippine traditional dances as basis on strengthening the love for culture and
the arts. American Journal of Arts and Human Science, 1(1), 27-33.
https://doi.org/10.54536/ajahs.v1i1
Diaz Rodriguez, J. M., & Patrick, D. (2023). Dancing national identity: The evolution of
meta-narratives in Colombian and Filipino folk dance. Retrieved from
https://researchcommons.waikato.ac.nz/items/310d2f75-e1bf-44fe-936a-1aff6f3b2209
Patrick, D. (2021). National identity in Philippine folk dance: Changing focus from the Cariñosa
to tinikling. In C. Parfitt (Ed.), Cultural memory and popular dance. Palgrave Macmillan.
https://doi.org/10.1007/978-3-030-71083-5_1
Pistola, T. (2021). Creating immersive experiences based on intangible cultural heritage: IEEE
Conference Publication | IEEE Xplore. Retrieved from
https://ieeexplore.ieee.org/abstract/document/9480948?fbclid=IwZXh0bgNhZW0CMTAAAR2z-
BxSj5_I2UFAh49pl8mHp1BjwxUWBR5OTlzrAswMZ_rq2B3LMAmNK64_aem_-7-Fr
45TIL3ErTPBPBTFfQ
Mariano, K.-L. (2024). LERION: Marikina's folk dance. Retrieved from
https://marikina.weebly.com/lerion-folk-dance.html
Zhang, Z., Tseng, & Weng. (2022). The role of digital museums in the dissemination of historical
dance traditions among young people. Retrieved from
https://l.facebook.com/l.php?u=https%3A%2F%2Fwww.ijih.org%2Fretrieve%2Fvolume
Dtl%2F1149%3Ffbclid%3DIwZXh0bgNhZW0CMTAAAR1sgm-k41fygoXUAnFUhLt
MD7MNv3wXyUPAnXYCyD11gmh1e3SBfX7maP4_aem_sO5IJLoa0WnFgheS0lQkdw
&h=AT0rPkv-3uXUe248QcjCkwzoHEb4rK6grbXFHQghxtLchikrVQNIwG9zCQbSR8
OTD9lpp91WDzMb22gEIrY9-8aZ1OU6RUy971XHYXtjj3wqlpqrPBXX0O2CYRcUuT
Eyv7SuhA

VR - MALAZA - MILLAN - SAZON - Santiago - Chapter 1-3

Uploaded by

Copyright:

Available Formats

VR - MALAZA - MILLAN - SAZON - Santiago - Chapter 1-3

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

VR - MALAZA - MILLAN - SAZON - Santiago - Chapter 1-3

Uploaded by

Copyright:

Available Formats

Kayumanggi: An Immersive Virtual Reality Journey into the Lerion and Cariñosa Folk

Dances of High School Students in Marikina

Almilyn Mae C. Malaza, Arjet R. Millan, and John Joshua B. Sazon

College of Computer Studies, Technological Institute of the Philippines - Quezon City

IT 304: Capstone Project 1

Ms. Nila D. Santiago

Figure 1 Example of Mature Dragon Fruit 14

Table 1 Summary of Constraints 39

profoundly impact Filipino communities and others around the world.

Cultural heritage encompasses more than physical monuments and objects; it

to create an immersive and authentic experience, enhancing understanding and appreciation

of these dances and ensuring their preservation for future generations.

Background of the Study

pervasive influence of modern entertainment forms pose significant challenges to the

transmission of these dances to younger generations. Innovative techniques are needed in

the unique identity of a community.

12 Program, designed to incorporate dancing into physical education subjects in the

national identity through its folk dances.

traditional dance instruction. According to Lobo, implementing innovative teaching

treasured by future generations if we cultivate a deeper appreciation and engagement.

Two individuals, attired in 19th-century Fil-Hispanic attire, exchange glances

of encouragement and reserved demeanor. The Cariñosa is an integral component of the

and cultural heritage (Diaz Rodriguez & Patrick, 2023).

According to Mariano (2024), Lerion, also known as Leron or Leron Sinta, is

surroundings, emphasizing a cultural connection where love for nature is expressed

The application provides a valuable opportunity for the preservation and

By allowing users to immerse themselves in a simulated environment and observe, study,

improves comprehension and appreciation of these dances, guaranteeing their accessibility

and preservation for future cultural enjoyment and education.

through cutting-edge VR technology.

Specifically, this project aims to:

● Developing a virtual reality application that simulates an immersive environment for

cultural appreciation in a virtual setting.

Cariñosa folk dance.

reliability, usability, performance efficiency, compatibility, maintainability, and

Significance of the Study

● Cultural Practitioners and Historians - This study provides an innovative tool to

accurately documented and can be passed down to future generations.

dances. It improves the educational experience by creating an immersive

● Future Researchers/Developers -This study will provide future researchers with a

gathering crucial information for future endeavors.

● Information and Communication Technology - The study exemplifies the successful

application of VR technology in cultural preservation, offering ICT professionals insights

into digital technology implementation in cultural contexts. It serves as an important

efforts, promoting the sustainable management of cultural heritage.

Scope and Delimitation

while its delimitations specify what is not included.

Scope of the Study

with the following:

○ This option allows users to enter the immersive virtual environments

using advanced motion capture technology.

○ Provides instructions and guidelines on how to use and navigate through

➢ Introduction to Controls: Explains how to use VR controllers to

interact with the virtual environment, such as executing dance

steps and navigating menus.

➢ Objective Explanation: Clarifies the learning outcomes related to

mastering dance techniques, understanding cultural contexts, and

completing interactive tutorials.

➢ Progression Explanation: Guides users on advancing through the