11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Allertify: Android-based Mobile Application for Food
Allergens Identification Using Barcode Scanning and Image
Recognition
Eniel Leba1, Jimea Aziel Faigao1, Mary Valen Magante1, Vernon Joseph Yeremia Tamba1, Ivar
Nelson Eñano Bingcang2*
Adventist University of the Philippines
INEBingcang@aup.edu.ph

ABSTRACT
This study introduces "Allertify," an Android-based mobile application designed to assist
individuals with food allergies by identifying potential allergens in food products through barcode
scanning and image recognition. The primary objective was to develop a comprehensive, userfriendly tool to minimize allergic reactions by providing quick and accurate allergen detection.
The need for a practical solution to help people make safer food choices, particularly in
environments where allergen exposure is a health risk, motivated the project. Allertify was
developed using Thunkable for front-end design and Google Teachable Machine for training the
image recognition model with a custom dataset. Barcode scanning and image recognition were
implemented to identify allergens in food products. The application was tested within a private
university campus, focusing on food products available in the cafeteria and store. The findings
showed that barcode scanning performed with a 100% success rate, with an average processing
time of 4 seconds. In contrast, the image recognition feature had a 91% success rate but only
achieved 50% accuracy, with an average processing time of 17 seconds. The lower accuracy for
image recognition was primarily due to a limited training dataset and visual similarities between
certain food items. The key limitations include a small food item database and the requirement for
an internet connection, which may restrict the application's availability in some settings. Despite
these challenges, Allertify demonstrates significant potential as a tool for improving food safety,
offering users real-time feedback on allergens. Future development could focus on expanding the
database, improving image recognition accuracy, and adding offline capabilities, making the
application more robust and reliable.
Keywords:
Food Allergens, Barcode scanning, Image recognition, Thunkable, Google
Teachable Machine

1542

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

INTRODUCTION
Food allergies affect millions of people worldwide, posing serious health risks ranging
from mild reactions to life-threatening conditions. These allergic reactions can be triggered by
even small amounts of allergens present in food, leading to symptoms such as swelling, vomiting,
difficulty breathing, and, in severe cases, anaphylaxis. In the Philippines, a food allergy reaction
sends someone to the emergency room every three minutes, highlighting the severity of the issue
(Gaas, 2021). While food poisoning often garners more public attention, food allergies present an
equally urgent but often overlooked public health concern. Current measures for managing food
allergies, such as stringent food labeling and consumer education, provide some level of protection,
but they are not foolproof. Unintentional allergen consumption remains a significant risk,
particularly for those with severe allergies, despite the efforts of food safety organizations and
health agencies (Grayson, 2022).
To address this challenge, various mobile applications have been developed to help
individuals with food allergies identify potential allergens in their food. Most of these solutions
rely on barcode scanning and databases to provide information about food ingredients. However,
a 2020 study reviewing 1,376 food allergy/intolerance-related applications revealed that only 14
were relevant and useful for users. These apps often lack reliability and comprehensiveness,
focusing on narrow functionalities, such as identifying allergens in restaurant menus or packaged
foods (Mandracchia, Llauradó, Tarro, Valls, & Solà, 2020). Barcode-based apps depend heavily
on the availability and accuracy of product barcodes and databases, which can be limiting when it
comes to fresh, unpackaged, or homemade food, where no barcode is present. Additionally, these
apps tend to lack personalized features, such as the ability to create and save custom allergen lists
based on the user's specific allergies (Mandracchia et al., 2020; Bollinger, 2020).
Given the increasing number of food allergy-related hospitalizations, there is a clear need
for more advanced, integrated, and user-friendly tools that combine multiple functionalities to
provide accurate and comprehensive allergen detection. Existing solutions fail to adequately
address the broad range of food environments, leaving individuals vulnerable, especially when
consuming unpackaged or freshly prepared foods (Bollinger, 2020). This study was motivated by
the need to enhance allergen identification tools by integrating two technologies—barcode
scanning and image recognition—to create a more comprehensive and versatile system for
detecting potential allergens.
The combination of barcode scanning and image recognition provides a dual-layered
approach to allergen detection. While barcode scanning works well for packaged foods, it is not
applicable to many real-life scenarios, such as eating at restaurants or consuming homemade meals
where barcodes are absent. Image recognition helps bridge this gap by allowing users to take
photos of their food for allergen identification, making the solution more robust (Chen et al., 2020).
By combining these two technologies, the application can address a wider range of food
environments, increasing its usefulness and reliability in helping individuals avoid allergens.
The primary objective of this study is to design and develop a mobile application that can
accurately and efficiently detect food allergens using both barcode scanning and image
recognition. "Allertify" is intended to give users a seamless experience for identifying potential
allergens in food items, thereby reducing the risk of allergic reactions in a wide variety of settings.
1543

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

The application was tested in a controlled environment in a private university in the Philippines,
specifically focusing on food items available in the cafeteria and store.
The application was developed using Thunkable for front-end design and Google
Teachable Machine for training the image recognition model with a custom dataset of food items.
The barcode scanning functionality retrieves product information, while the image recognition
function identifies food items directly from user-captured images. The combination of these two
technologies enables a more comprehensive and personalized approach to allergen identification,
as it allows users to check both packaged and unpackaged foods (Chen et al., 2020; Herrmann et
al., 2017).
The results of this study show that Allertify effectively identifies potential allergens by
comparing food ingredients with a user-defined allergen list. However, there are limitations,
including the need for an internet connection and a relatively small food item database, which
limits the system’s coverage. Despite these challenges, Allertify demonstrates significant potential
as a valuable tool for improving food safety. Future development could focus on expanding the
database, improving image recognition accuracy, and adding offline capabilities to enhance
usability in various environments. This study contributes to the fields of health and food safety by
providing a practical, innovative solution for managing food allergies, with the potential to
significantly improve the quality of life for individuals with food sensitivities.

LITERATURE REVIEW
The literature review delves into the technological advancements and methodologies in
food allergen identification, focusing on image recognition, barcode scanning, and the
development of mobile applications.
Technical Background
The advent of image recognition and computer vision marked significant milestones in
automated image analysis, with deep learning algorithms playing a crucial role. Computer vision,
emerging in the 1960s, aimed to replicate human vision systems to automate the process of image
analysis. Deep learning has enhanced the capability of computers to identify and categorize images
with a vast dataset, employing benchmarks to compare new images for more precise recognition
(Pulsar Platform, 2018)
Barcode scanning technology has evolved from the 1950s and has become a prevalent
method for product identification. Initially used in grocery stores, the Universal Product Code
(UPC) became a standard for tracking product information. While barcode scanning offers a costeffective means of product tracking, it has limitations such as line-of-sight dependency and issues
with improperly printed barcodes (Damle et al., 2020; Elaskari et al., 2021; Thanapal et al., 2017)
Review of Related Systems

1544

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Foreign systems leveraging barcode scanning and image recognition have shown promise
in enhancing dietary tracking and food identification. Ming et al. (2018) developed DietLens, an
app utilizing a deep learning algorithm for food recognition, allowing users to identify ingredients
in food images and obtain nutritional information. Similarly, Chaturvedi et al. (2020) created a
mobile application using deep learning to estimate nutritional content based on images of food
items, thus raising awareness about food consumption and dietary needs.
Local systems have also explored image recognition in different domains. For instance, De
Goma & Devaraj (2020) developed a system using image processing and machine learning to
categorize skin conditions, demonstrating the potential of image recognition in healthcare
applications.
Integration of Thunkable and Google Teachable Machine
The integration of machine learning tools such as Google Teachable Machine (GTM) with
application development platforms like Thunkable has facilitated the creation of user-friendly
interfaces and efficient machine learning models. GTM employs convolutional neural networks
(CNNs) for data analysis and offers different ways to train models. It utilizes transfer learning,
which involves taking knowledge from one domain and applying it to another, enhancing the
efficiency of model training (Prasad et al., 2022. Several studies have utilized GTM for various
recognition tasks, including facial recognition for fraud detection (Shoukat & Akram, 2021) and
identification of plant seeds with high accuracy (Chazar & Rafsanjani, 2022)
Conceptual Framework
The conceptual framework of the Allertify application illustrates the workflow and data flow
between the end user and the internal systems of the application. This framework is designed to
showcase the input-output process and provide a detailed structure of how the application
functions. The main aim is to depict the steps involved in allergen identification using barcode

1545

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

scanning and image recognition, providing an understanding of how the application processes user
inputs to generate outputs.
Figure 1-Conceptual Framework

1. Data Input and Storage:
o The system’s database is initially populated with a list of common food allergens
and food items obtained from the Adventist University of the Philippines (AUP)
store. Each food item in the database is associated with its ingredients, name, and
barcode value.
o For image recognition, the researchers train the model using Google Teachable
Machine (GTM) by inputting image samples of food items served in the AUP
cafeteria. This trained model is uploaded online to be accessible by the application.
The food items' names and ingredients are stored in the database.
2. User Interaction:
o Users enter their allergen list, which is stored locally on the device. They can select
allergens from a predefined list of common allergens available in the system’s
database or manually input their own allergens.
o Users can then choose to scan food items using either barcode scanning or image
recognition. The system will process the uploaded image or barcode to identify the
food item.
3. Food Item Identification:
o If the food item is not in the database, the application displays a "not found" result
to the user.
o If the food item is found in the database, the system identifies potential allergens
by comparing the food item's ingredients list with the user's allergen list. The
application will indicate whether the food item is potentially safe or unsafe for the
user, based on this comparison.
4. Output and History:
o The application adds the result to a history list, stored locally, allowing users to
keep a record of their scans. The result is then displayed on the user's device,
providing real-time feedback on the safety of the food item.

METHODS
The methodology for the Allertify project is centered on developing a mobile application using a
combination of machine learning, image recognition, and agile development practices. The
approach involves utilizing Google Teachable Machine (GTM) for image recognition and
Thunkable for front-end development, with Node.js and MySQL for backend processes (Prasad et
al., 2022).

Machine Learning and Google Teachable Machine

1546

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

The project utilizes machine learning through Google Teachable Machine (GTM), an
accessible platform designed for easy model training. GTM was selected for this project due to its
ability to build image recognition models quickly and efficiently without requiring extensive
coding knowledge. The platform leverages Convolutional Neural Networks (CNNs) to classify
data, which makes it suitable for tasks such as food item identification, a key feature of Allertify.
In this study, GTM was employed to develop the image recognition model, which is crucial
for identifying food items based on user-uploaded images. The training process began with the
collection of image data, which was uploaded by the developers. These images were then
organized into specific categories corresponding to different food items. GTM uses these
categories to train the model.
The model training process followed a standard machine learning pipeline, where the data
was divided into two sets: 85% of the data was used for training, and 15% was reserved for testing
the model's performance. This approach ensured that the model was adequately trained while also
allowing for validation during testing. GTM's interface simplifies this process, allowing users to
observe how well the model performs in real-time based on the provided dataset.
A total of 23,649 pictures, separated into seven classes, were used to train and test the
model. The number of the training samples is shown in

Figure 2. Classes and Number of Training Samples

1547

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

As mentioned GTM uses 85% of the samples provided for training and 15% for testing. It
then provided data to the researchers on the accuracy of the model. According to the internal
testing, the model has 100% accuracy for all the different classes (Figure 3) and there is no
confusion between classes (Figure 4.).

Figure 3. Number of testing
samples and accuracy per class

Figure 4. Confusion Matrix Model

GTM leverages the TensorFlow.js library, a powerful JavaScript library for machine
learning, which allows the trained model to be deployed seamlessly within a web environment or
integrated into mobile applications like Allertify. TensorFlow.js enables the model to run
efficiently on client devices, ensuring smooth real-time image recognition without the need for
extensive server-side resources (Prasad et al., 2022). The combination of GTM's CNN architecture
and TensorFlow.js library makes it an ideal choice for implementing image recognition in the
Allertify app.
Figure 5 in the document outlines the step-by-step process used to train the model in Google
Teachable Machine.

Figure 5. Step by Step Process used to ttrain the model

Thunkable
Thunkable was selected for the Allertify project due to its ease of use and no-code platform, which
allowed the development team to focus on building core functionalities like barcode scanning and
1548

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

image recognition without extensive coding knowledge (Thunkable, 2023). Its drag-and-drop
interface made it possible to quickly prototype and iterate the app, which helped enhance the user
experience and streamline the development process (Thunkable, 2023).
In addition, Thunkable’s cross-platform compatibility allowed for simultaneous development for
both Android and iOS, a key factor for the project’s scalability (Thunkable, 2023). Built-in
features, such as camera access and barcode scanning, further simplified the development process,
making Thunkable a more efficient choice compared to traditional development platforms like
Android Studio, which require extensive coding (Thunkable, 2023).
When compared to other platforms like Flutter and AppGyver, Thunkable provided the right
balance between functionality and ease of use. While other platforms may offer more advanced
options, they often require greater technical expertise, which can slow down development and
testing (Thunkable, 2023).

Figure 6. Example of Thunkable Code Blocks

1549

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Agile Methodology and Scrum
The development process followed the Agile methodology, specifically utilizing the Scrum
framework. Agile was chosen due to its iterative and incremental nature, which is well-suited for
mobile application development. The Scrum method involved planning user stories, constructing
a product backlog, and implementing a sprint plan to guide development. The team organized the
project into three sprints, each focusing on different aspects of the application (Al-Saqqa, Sawalha,
& Nabi, 2020; Cerna et al., 2018; Hassan et al., 2022).

System Design and Architectural Design
The system was designed using a three-tier architecture consisting of the presentation tier (user
interface), application tier (processing of data), and data/database tier (storing of data). Thunkable
was used for front-end development to create the user interface, while Node.js was employed to
handle data processing between the front end and back end. Data storage was managed using
MySQL and Thunkable Data Source databases. The user interface was built to facilitate smooth
interaction and data processing (Kendall & Kendall, 2012. Figure 2 in the document provides a
visual representation of the application’s architecture.

Figure 7. Architectural Design

1550

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Development and Testing
The development process involved simultaneous front-end and back-end development,
utilizing Thunkable for building the application’s UI and functionalities. The developers tested the
front end using Thunkable's "live test on device" feature, while the back end was tested using
Postman to validate the API. The testing phase also included user testing with at least ten users to
provide feedback on the app's accuracy, quality, and functionality
In summary, the methodology involved leveraging machine learning for image recognition
using GTM, adopting Agile Scrum practices for project management, and implementing a threetier system architecture to build a functional and reliable mobile application for allergen
identification.
Results and Discussion
The Allertify project aimed to provide an effective tool for food allergen identification
using barcode scanning and image recognition. The testing phase involved evaluating the
application’s performance across these two core features, assessing their accuracy, efficiency, and
user satisfaction.

Overall Test Summary
During the testing phase, a total of 65 scans were conducted to assess the application's
performance. Of these scans, 62 were successful, indicating that the system was generally robust
in processing user inputs. However, 3 scans were unsuccessful due to unfinished processing and
server errors, highlighting areas for improvement in system reliability and server handling. Among
the 62 successful scans, 48 produced correct outputs while 17 scans provided incorrect results,
suggesting variability in the application's effectiveness depending on the scanning method used.
Barcode Scanning
The barcode scanning feature exhibited a high degree of accuracy and reliability. Out of
31 tests performed using this feature, all scans were successful, resulting in a 100% success rate.
The average processing time for each barcode scan was approximately 4 seconds, indicating a
quick and efficient performance. This rapid processing time is crucial in real-world settings where
users may need immediate information about food products to make safe dietary choices. All
barcode scans returned correct outputs, affirming the feature's capability to accurately identify
food items and potential allergens through barcode data. The reliability of barcode scanning in
identifying food items is supported by research indicating that barcode systems can provide quick
and accurate information retrieval, making them a valuable tool in various industries, including
healthcare and food safety (Koutkias et al., 2013; Yang & Ho, 2017). The barcode scanning’s
success suggests that it is a highly reliable method for allergen identification when product
barcodes are available and correctly stored in the applications.

1551

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Table 2
Summary of barcode training test
Number of Scan
Successful Scan
Unsuccessful Scan
Average Process
Time (seconds)
Correct Output
Incorrect Output

31
31
0
4
31
0

Image Recognition
The image recognition feature showed more variability in its performance compared to
barcode scanning. A total of 34 image recognition tests were conducted, with 31 scans being
successful, resulting in a 91% success rate. However, the accuracy of the image recognition feature
was lower than that of barcode scanning, with only 17 out of 34 scans (50%) yielding correct
outputs. The average processing time for image recognition was 17 seconds, which is considerably
longer than barcode scanning, reflecting the complexity involved in processing and analyzing
images. Notably, the "Food not Recognized" error occurred in 7 instances, primarily due to the
limited variety of image samples used during the training phase and the visual similarities among
certain food items. These findings indicate that while image recognition has potential as an allergen
identification tool, its effectiveness is currently constrained by the limitations of the training
dataset and the inherent challenges of accurately identifying visually similar food items. Similar
challenges in food image recognition have been noted in other studies, where the need for large
and diverse training datasets is emphasized to improve model accuracy and robustness (Chen et
al., 2020; Herrmann et al., 2017).
Table 2
Summary of image recognition tests
Number of Scan
Successful Scan
Unsuccessful
Scan
Average Process
Time (seconds)
Correct Output
Incorrect Output
Food not
Recognized
Server Error

34
31
3
17
17
17
7
1

1552

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Unfinished
Process

1

User Satisfaction
To assess user perception of the application's utility and user-friendliness, a user
satisfaction survey was conducted with nine participants. The survey aimed to evaluate various
aspects of the application, including ease of use, navigation, effectiveness, and overall usefulness.
The results were positive, with most participants rating the app favorably. Specifically, the average
response to the question regarding the app's usefulness for food-allergic individuals was 4.9 out of
5. This high score suggests that users found the application valuable and effective in aiding them
with allergen identification. Users also rated the app highly in terms of ease of use and navigation,
indicating that the application's design successfully provided a user-friendly experience. Previous
research has highlighted the importance of user-friendly interfaces in mobile health applications,
as ease of use significantly influences user engagement and the overall effectiveness of healthrelated tools (Dennison et al., 2013).
CONCLUSION, IMPLICATION, SUGGESTION, AND LIMITATIONS
The Allertify app was developed as a user-friendly mobile solution to assist individuals
with food allergies in identifying potential allergens and making safe food choices. The project
successfully achieved most of its objectives, with the application featuring two primary
functionalities: barcode scanning and image recognition. Users are able to input personalized
allergen lists and select their preferred scanning method, after which the app processes the
information and determines whether the scanned food item is safe for consumption. The app
operates optimally with an internet connection and can be installed through an APK file, as it was
developed using the free version of Thunkable and is not yet available on the Google Play Store.
The barcode scanning feature performed reliably, accurately reading product barcodes and
providing immediate results. This aligns with existing research showing that barcode scanning is
a highly dependable method for identifying food products and ingredients, aiding in allergen
detection (Koutkias et al., 2013; Yang & Ho, 2017). However, the image recognition feature, while
functional, only partially met its objective, with an accuracy rate of 50%. This limitation was
attributed to the methods used in capturing food images and the visual similarities among food
items in the dataset. These challenges are documented in other studies, which highlight the
importance of diverse datasets for improving accuracy in image recognition models (Chen et al.,
2020; Herrmann et al., 2017). Despite these challenges, the image recognition feature has the
potential for significant improvement through further training with more varied food images,
different angles, and lighting conditions.

Implications
The Allertify app has broader implications for food safety, as it provides an accessible and
practical tool for individuals with food allergies. By integrating both barcode scanning and image
1553

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

recognition, the application offers real-time allergen detection, empowering users to make
informed dietary decisions. This project aligns with ongoing efforts to leverage mobile health
technologies to enhance public health outcomes, particularly in managing food allergies and
preventing allergic reactions (Mandracchia et al., 2020). The combination of these technologies in
daily life contributes to a growing body of evidence supporting the use of mobile health tools for
improving health and safety, offering a more holistic solution compared to existing apps.
The scalability of Allertify also holds promise for adoption in various settings beyond
personal use, such as restaurants, grocery stores, and other food establishments. Barcode scanning
can be extended to packaged foods in supermarkets, while image recognition can analyze dishes
in restaurants where allergen labeling may not be readily available. This scalability is critical in
broadening the impact of Allertify, making it applicable across different food environments.

Suggestions for Future Development
For future improvements, the project suggests several enhancements. First, expanding the
variety of food items in the application will make it more comprehensive and interactive. This
includes training the model by capturing images from different angles, lighting conditions, and
placing food items on various objects such as plates, tables, and bowls. Enhancing the barcode
feature by increasing the number of detectable items would also be beneficial. The importance of
diverse and comprehensive training datasets is supported by previous research, indicating that such
improvements can significantly enhance the accuracy and robustness of image recognition models
(Chen et al., 2020). Second, implementing a PIN or password lock for the application would
improve user security, aligning with best practices in mobile application development to ensure
data privacy and user protection (Dennison et al., 2013). Third, allowing users to save their data
in the cloud by creating an account would enable data transfer between devices, enhancing the user
experience. Implementing these recommendations would result in a more robust and sophisticated
application.

Limitations
Several limitations were identified during the development of the Allertify application. One
significant limitation is the limited number of food items available for both barcode and image
recognition. For barcode scanning, the database included only six food items due to the
unavailability of additional data because of confidentiality concerns. For image recognition, the
model's accuracy was constrained by the limited variety of food items available in the cafeteria.
Only 83 dishes were captured and categorized into six classes, limiting the number of food items
and products the application could recognize. Similar limitations in food recognition applications
have been noted, where limited training datasets result in lower accuracy rates (Herrmann et al.,
2017). Another limitation is the requirement for an internet connection to access the application,
which could always restrict its availability to users. This dependency highlights the need for offline
capabilities in mobile health applications to ensure broader accessibility (Dennison et al., 2013).
The mentioned limitations serve as areas for improvement and future development, ensuring that

1554

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

future versions of Allertify or similar applications can offer a more comprehensive and reliable
solution for food allergen identification and management.
ACKNOWLEDGEMENT
The developers of the Allertify project expressed their heartfelt gratitude to several
individuals and institutions that supported them throughout the project's journey. First and
foremost, they thanked God for granting them knowledge, intelligence, and strength to complete
the study. They also conveyed their sincere gratitude to their capstone advisor for the guidance,
time, and valuable feedback provided during the project's development. Appreciation was
extended to their college professors for the knowledge and support they offered during the study.
Additionally, the developers acknowledged their parents for their unwavering support from
the project's inception to its conclusion. Lastly, they expressed their deepest thanks to the
supervisors and staff of the Adventist University of the Philippines cafeteria for their patience and
understanding during the project's development, noting that the project would not have been
possible without their cooperation.

REFERENCES
Bollinger, M. E. (2020). Mobile Applications for Food Allergy: Enhancing Safety and
Knowledge. Journal of Allergy and Clinical Immunology: In Practice, 8(2), 558-565.
Chen, J., Jia, J., & Chen, B. (2020). Food Image Recognition via Integrated Learning Approach.
IEEE Access, 8, 69839-69850.
Damle, A., Bangera, M., Tripathi, S., & Meena, M. (2020). Analysis of Barcode Scanning and
Management. AMRIDDHI: A Journal of Physical Sciences, Engineering and Technology,
12(SUP 1), 90-95.
Dennison, L., Morrison, L., Conway, G., & Yardley, L. (2013). Opportunities and Challenges for
Smartphone Applications in Supporting Health Behavior Change: Qualitative Study.
Journal of Medical Internet Research, 15(4), e86.
Dennison, L., Morrison, L., Conway, G., & Yardley, L. (2013). Opportunities and Challenges for
Smartphone Applications in Supporting Health Behavior Change: Qualitative Study.
Journal of Medical Internet Research, 15(4), e86.

1555

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Elaskari, S., Imran, M., Elaskri, A., & Almasoudi, A. (2021). Using Barcode To Track Student
Attendance And Assets In Higher Education Institutions. Procedia Computer Science, 184,
226–233. Obtenido de https://pdf.sciencedirectassets.com/280203/1-s2.0S1877050921X00075/1-s2.0-S187705092100778X/main.pdf?X-Amz-SecurityToken=IQoJb3JpZ2luX2VjEL3%2F%2F%2F%2F%2F%2F%2F%2F%2F%2FwEaCXVzL
WVhc3QtMSJHMEUCIQCds%2BtWkQQQ5LArV%2BHhNlWMJXiDISi%2FRK5U2Qx
dFbGVWgIgLePKJI
Gaas, M. A. (06 de July de 2021). Serious Facts You Need To Know About Food Allergies.
Obtenido de National Nutrition Council: https://nnc.gov.ph/regionaloffices/mindanao/region-ix-zamboanga-peninsula/5597-serious-facts-you-need-to-knowabout-food-allergies
Grayson, M. (April de 2022). Allergy Facts and Figures. Obtenido de Asthma and Allergy
Foundation of America: https://aafa.org/allergies/allergy-facts/
Herrmann, M., Wanjek, F., & Patil, K. R. (2017). Towards Machine Learning in Food
Informatics: A Review on the Structural Learning of Food Graphs. Frontiers in Artificial
Intelligence and Applications, 301, 315-320.
Koutkias, V. G., Lillo-Le Louet, A., & Jaulent, M. C. (2013). Exploiting the Potential of Barcode
Technology in Healthcare: Current Research and Future Trends. International Journal of
Medical Informatics, 82(11), 1116-1128.
Koutkias, V. G., Lillo-Le Louet, A., & Jaulent, M. C. (2013). Exploiting the Potential of Barcode
Technology in Healthcare: Current Research and Future Trends. International Journal of
Medical Informatics, 82(11), 1116-1128.
Mandracchia, F., Llauradó, E., Tarro, L., Valls, R. M., & Solà, R. (2020, August 08). Mobile
Phone Apps for Food Allergies or Intolerances in App Stores: Systematic Search and
Quality Assessment Using the Mobile App Rating Scale (MARS). JMIR mHealth and
uHealth.
Pulsar Platform. (15 de August de 2018). A brief history of Computer Vision and AI Image
Recognition. Obtenido de Pulsar: https://www.pulsarplatform.com/blog/2018/brief-historycomputer-vision-vertical-ai-image-recognition/
Thanapal, P., Prabhu, J., & Jakhar, M. (2017). A survey on barcode RFID and NFC. IOP Conf.
Series: Materials Science and Engineering, 263(4), 042049.
Thunkable. (2023). Thunkable: No-code mobile app development platform. Retrieved from
https://thunkable.com

1556

11th ISC 2024 (Universitas Advent Indonesia, Indonesia)
“Research and Education Sustainability: Unlocking Opportunities in Shaping Today's
Generation Decision Making and Building Connections” October 22-23, 2024

Yang, Y., & Ho, C. C. (2017). Barcode Technology Applications in Healthcare: An Overview.
Health Information Science and Systems, 5(1), 1-8.

1557