International Journal of Artificial Intelligence p-ISSN: 2407-7275, e-ISSN: 2686-3251 Original Research Paper Development of a Student Expense Tracking System Using Optical Character Recognition Ahmad Fadli Saad1*.
Muhammad Hairil Shaharudin1.
Achmad Yani2.
Abdi Manaf2 Computing Science Studies.
College of Computing.
Informatics and Media.
Universiti Teknologi MARA.
Perak Branch.
Tapah Campus.
Perak.
Malaysia.
Department of Industrial Engineering.
Sekolah Tinggi Teknik Ar-Rahmah.
Bintan.
Indonesia.
Article History Received:
Revised:
Accepted:
*Corresponding Author:
Ahmad Fadli Saad Email afadlis@uitm.
This is an open access article, licensed under: CCAeBY-SA Abstract: Personal financial literacy is a vital skill for university students, yet many struggle to track their daily expenses due to time constraints and low This study aims to design and develop a web-based Student Expense Tracking System using Optical Character Recognition (OCR) technology to address this issue.
The system allows users to automatically extract and record spending information from receipt images, reducing manual input and improving financial awareness.
The development followed the Web Development Life Cycle (WDLC) using the Waterfall model, comprising planning, design, development, and testing phases.
Visual Studio Code.
Python 3, and Tesseract OCR were employed in system Wireframes and mockups guided the interface design, while backend development focused on data storage and OCR integration.
Functionality testing showed a 100% pass rate across ten scenarios, validating the system's performance in image processing, budget management, and spending visualization.
Usability testing using the Post-Study System Usability Questionnaire (PSSUQ) with 30 participants yielded a mean score 45 out of 5, indicating a high level of user satisfaction.
The system scored highest on ease of use .
, visual design .
, and recommendation likelihood .
, confirming its intuitive interface and appeal.
Slightly lower scores in user confidence .
and data organization .
point to opportunities for interface refinement and improved user guidance.
This research concludes that OCR can effectively support financial tracking for Future enhancements with NLP and machine learning are recommended to automate expense categorization and improve analytical Keywords: Budget Management.
Financial Literacy.
OCR Technology.
Student Expense Tracking.
Usability Testing.
2025 | International Journal of Artificial Intelligence | Volume.
12 | Issue.
1 | 1-10
Ahmad Fadli Saad.
Muhammad Hairil Shaharudin.
Achmad Yani.
Abdi Manaf.
Development of a Student Expense Tracking System Using Optical Character Recognition.
International Journal of Artificial Intelligence, vol.
12, no.
1, pp.
June 2025.
DOI: 10.
36079/lamintang.
Introduction Personal financial management since college is an important aspect in the formation of financial Students generally start to be responsible for their own income and expenses, so it is important to develop the habit of recording and managing finances from an early age.
This habit not only helps maintain daily financial stability but also forms the foundation of healthy financial management in the future .
The reality in the field shows that students often have difficulty in recording their finances Academic activities, organizational activities, and a busy social life often reduce the time and motivation to record expenses manually.
As a result, many of them lose track of their expenses and are at risk of wasting or even running out of funds at the end of the month .
In addition to time constraints, low financial literacy is also one of the main factors causing unwise financial decisions among students.
Lack of understanding of the importance of budget planning makes them tend to spend money on less priority needs.
This can lead to ongoing personal financial instability .
As a solution to this problem, technology can be used to automate the process of recording One relevant technology is Optical Character Recognition (OCR), which allows automatic data extraction from images of shopping receipts.
By simply photographing the receipt, the system can recognize and store transaction information, so that users do not need to record it manually .
Several studies have shown that the use of OCR in financial applications provides high efficiency and a reliable level of accuracy.
A study found that an OCR-based tracking system was able to identify and categorize transactions with high precision, and increase user satisfaction by up to 86.
This technology has also been shown to be compatible with machine learning systems for more comprehensive analysis and visualization of personal financial data .
This study aims to design and develop a web-based expense tracking system using OCR The system is designed to be easily used by students with features such as automatic recording of shopping receipts, daily spending suggestions, budget visualization graphs, and notifications when the budget is exceeded.
This application is built using Python through the Visual Studio Code platform .
With this system, it is hoped that students can be wiser in managing their finances, understand their spending patterns, and increase awareness of the importance of financial literacy from an early age.
In addition to providing practical solutions to student problems, this research also provides technical contributions to the development of OCR applications in the context of personal finance in the digital era .
Literature Review Student Financial Behavior Students' financial behavior is an important focus because they are in the transition stage towards financial independence.
Financial literacy plays a major role in forming healthy financial habits, such as saving and wise spending.
According to Nurwulandari .
, students with adequate financial understanding tend to be able to distinguish between needs and wants, so they are more responsive to financial risks.
Falahati and Paim .
added that awareness of money management instruments such as budgets, savings, and investments can increase students' economic resilience.
Asandimitra and Kautsar study .
showed that students who received basic financial education had better spending control and were not easily influenced by a consumptive lifestyle.
Similar findings were put forward by Sabri .
that early literacy skills improve the prospects for long-term financial well-being.
Saving Students' saving habits are influenced by a number of internal and external factors.
Hamzah et al.
stated that cultural values and family habits in managing money are the initial factors in the formation of saving behavior.
Abdul Kadir et al.
showed that peer pressure often encourages students to spend money for social acceptance.
Kumar et al.
's .
study indicated that regular parental guidance can encourage students to set aside money consistently.
OECD .
found that financial applications with automatic reminders increased students' commitment to saving in a disciplined manner.
Expenditures In terms of expenditures, students generally face challenges in managing daily spending due to the influence of impulsiveness and lack of budgeting skills.
Manju .
stated that although many Ahmad Fadli Saad.
Muhammad Hairil Shaharudin.
Achmad Yani.
Abdi Manaf.
Development of a Student Expense Tracking System Using Optical Character Recognition.
International Journal of Artificial Intelligence, vol.
12, no.
1, pp.
June 2025.
DOI: 10.
36079/lamintang.
students intend to control spending, only a few implement budgets consistently.
Adam et al.
linked the lack of financial control to emotional distress and academic stress.
Bialowolski et al.
noted that students who do not record their finances are more susceptible to small debts that accumulate, leading to long-term financial instability.
Image Processing Technology for Financial Recording Image processing technology now plays a role in the automation of financial recording, including capturing and recognizing information on student shopping receipts.
Image processing is used to enhance image quality, detect objects (OCR), segment text, and perform image restoration .
Image Enhancement Enhancement is the initial stage in image processing.
Techniques such as histogram equalization and CLAHE are used to improve the contrast and lighting of the image.
Research by Venkatesh et.
, al .
showed that enhanced images improve OCR accuracy by up to 20%.
Frequency-based techniques have also proven to be efficient in improving visual details of faded receipts .
Object Detection and OCR Text and number detection on receipts is usually done with Optical Character Recognition (OCR).
Models such as YOLO and SSD are used for fast and real-time detection .
Ma and Doermann .
applied OCR to documents with non-standard structures and achieved 92% accuracy.
Singh and Priya .
introduced a hybrid OCR that still works in offline mode, suitable for students in areas with limited connectivity.
Image Segmentation Segmentation is needed to separate text from the background of the receipt.
Algorithms such as watershed, region growing, and adaptive thresholding are used for high-precision segmentation.
Han et al.
showed that a combination of color and contour-based segmentation results in more precise data extraction than traditional thresholding.
Image Restoration Restoration is used to clean noise or improve poor lighting.
Filters such as median and Wiener have been shown to improve image readability before OCR processing .
A study by Rashhed and Khan .
showed an image quality improvement of up to 35% in extreme lighting conditions using wavelet-based restoration techniques.
Popular Financial Applications and Supporting Features Various OCR-based financial management applications have been circulated, but their adoption among students is still low due to complex or unresponsive interfaces.
C Money Manager Expense & Budget: Supports receipt photo-based expense tracking and automatic category grouping .
C Monefy: Has a simple UI but does not support direct OCR features from the camera .
C My Cash Flow: Has a cloud backup system and supports spending grouping with weekly graphs .
Reasons for Selecting the Approach The OCR-based system is considered to be the most appropriate for students' needs because .
C Its ability to process physical receipts from various sources C It does not rely on time-consuming manual input C It can be integrated with data visualization and automatic budgeting Methodology The methodology plays a crucial role in ensuring the project runs smoothly and produces valid results aligned with the aims and objectives.
This chapter elaborates on the methods used to design, develop, and test the Student Expense Tracking System utilizing Optical Character Recognition (OCR).
The development process follows a structured approach incorporating specific phases, tasks, techniques.
Ahmad Fadli Saad.
Muhammad Hairil Shaharudin.
Achmad Yani.
Abdi Manaf.
Development of a Student Expense Tracking System Using Optical Character Recognition.
International Journal of Artificial Intelligence, vol.
12, no.
1, pp.
June 2025.
DOI: 10.
36079/lamintang.
and tools to ensure systematic progress.
The entire project methodology is divided into three main phases: system design, system development, and functionality testing, which collectively facilitate the successful creation of the expense tracking application.
The project employs the Web Development Life Cycle (WDLC) Waterfall model as the core development framework.
This model consists of five sequential phases: planning and analysis, design, development, testing, and maintenance.
In the initial phase, thorough planning and analysis were conducted to identify the systemAos key requirements, objectives, and scope.
Extensive research was performed to study similar applications, enabling the team to derive new ideas and improve upon existing systems.
The main function identified is to support students in tracking their expenses efficiently by scanning receipts via OCR, thereby reducing manual entry and enhancing accuracy.
In the design phase, wireframes and mockups were created to define the user interface, information architecture, and user flow clearly.
These visual prototypes guided the development team in building a system that meets user expectations in both functionality and usability.
The subsequent development phase involved implementing the systemAos frontend and backend components.
Frontend development focused on creating a seamless and intuitive user experience, while backend development managed data storage, server management, and integration with the OCR engine.
Tools such as Visual Studio Code.
Python 3, and Tesseract OCR were employed to facilitate the coding and image processing The testing phase aimed to verify the systemAos functionality, usability, and performance through iterative testing on a local host environment.
Bugs and errors identified during development were promptly addressed to ensure a high-quality final product.
Although maintenance is a vital phase in software development, it was excluded from this project due to scope limitations.
Overall, this methodology ensures a coherent and disciplined approach from system conception through to deployment, enabling the creation of an effective student expense tracking system powered by OCR Finding and Discussion The system is designed to help students record expenses easily through OCR technology.
The main components include a user interface, flowchart, and entity relationship diagram (ERD).
The database consists of three main tables: users, budget, and daily_expenses, with one-to-many relationships between them.
Figure 1.
ERD
The ERD illustrates the relationships between users, monthly budget, and daily expenses shows in Figure 1, while the Process Flowchart shows the system flow from user login to recording expenses using OCR shows in Figure 2.
Ahmad Fadli Saad.
Muhammad Hairil Shaharudin.
Achmad Yani.
Abdi Manaf.
Development of a Student Expense Tracking System Using Optical Character Recognition.
International Journal of Artificial Intelligence, vol.
12, no.
1, pp.
June 2025.
DOI: 10.
36079/lamintang.
Figure 2.
System Flow from User Login to Recording Expenses using OCR Algorithm: OCR Integration for Student Expense Tracking System Start the upload function.
Check if 'images' key exists in request.
o If not found, return a response: "No image uploaded!" with status code 400.
Retrieve the image file from request.
files['images'].
Load the image using PIL and assign it to the variable image.
Try performing OCR using pytesseract:
o a.
Extract text from the image using image_to_string.
, store it in ocr_result.
o b.
Define the keyword "TOTAL" as the target search string.
o c.
Search for the index of "TOTAL" in ocr_result and assign it to total_price_index.
o d.
If total_price_index != -1 .
eyword foun.
C i.
Slice the string after the "TOTAL" keyword to extract the total amount.
C ii.
Save the uploaded image to a predefined directory using the save() C i.
Redirect the user to the add_expense route, passing the extracted amount as a parameter .
Except: If any error occurs during OCR processing:
o Return an error response with the exception message and a status code 500.
The algorithm uses pytesseract to extract text from the image and search for the keyword AuTOTALAy.
The value after this keyword is taken as the amount of expenditure and sent to the add_expense page.
Ahmad Fadli Saad.
Muhammad Hairil Shaharudin.
Achmad Yani.
Abdi Manaf.
Development of a Student Expense Tracking System Using Optical Character Recognition.
International Journal of Artificial Intelligence, vol.
12, no.
1, pp.
June 2025.
DOI: 10.
36079/lamintang.
Figure 3.
OCR Integration for Student Expense Tracking System Discussion .
Overview The main findings from the development and evaluation of the Student Expense Tracking System using Optical Character Recognition (OCR).
The system was designed to aid university students in managing their expenses more effectively by automating the process of extracting financial data from The results from the functionality and usability testing presented in Chapter 4 are interpreted in this section, with a focus on system performance, user experience, and potential improvements.
Interpretation of Functionality Testing Results The functionality testing, as shown in Table 1, demonstrates that the system successfully met all specified requirements.
Each test scenario, from user authentication to receipt uploading.
OCR-based text extraction, expense tracking, and budget settingAiproduced expected results with a "Passed" This indicates that the system is technically sound and stable across its core features.
A critical function.
OCR text extraction, was successfully implemented, as indicated in test scenarios 4 to 6.
The accurate identification of total amounts from receipts and appropriate handling of extraction failures .
ith error message.
reflects the robustness of the OCR implementation.
This functionality is central to the systemAos value proposition and its effectiveness in automating expense Moreover, the ability to set budgets, track spending patterns, and manage expenses .
dd, update, delet.
highlights the comprehensiveness of the system as a financial management tool tailored for Table 1 shows the test cases for the main user interface components including Login.
Registration, and Navigation between pages.
Ahmad Fadli Saad.
Muhammad Hairil Shaharudin.
Achmad Yani.
Abdi Manaf.
Development of a Student Expense Tracking System Using Optical Character Recognition.
International Journal of Artificial Intelligence, vol.
12, no.
1, pp.
June 2025.
DOI: 10.
36079/lamintang.
Table 1.
Functionality Testing Result Test Scenario User login with valid User login with invalid User uploads a receipt OCR successfully extracts text from the Total amount is found in extracted text Expected Result User is successfully logged in User login fails with appropriate error Image is successfully uploaded and saved in the system Text is accurately extracted from the uploaded image Total amount is correctly identified and displayed Total amount not found in extracted text User adds a new User sets a monthly User receives an error message to re-upload the Expense is added and saved in the system Budget is set for the specified month and year User views spending User updates or deletes an expense Spending pattern is Expense is successfully updated or deleted Actual Result User is successfully logged in User login fails with appropriate error Image is successfully uploaded and saved in the system Text is accurately extracted from the uploaded image Total amount is correctly identified and User receives an error message to re-upload the image Expense is added and saved in the system Budget is set for the specified month and Spending pattern is Expense is successfully updated or deleted Status Passed Passed Passed Passed Passed Passed Passed Passed Passed Passed .
Interpretation of Usability Testing Results Usability testing using the Post-Study System Usability Questionnaire (PSSUQ) revealed generally positive feedback from the 30 participants.
The overall mean usability score was 4.
45 out of 5, suggesting a high level of satisfaction among users.
The highest scoring itemsAiease of use .
, visual design .
, and recommendation likelihood .
Aiindicate that students found the system intuitive and aesthetically pleasing.
These aspects are critical for adoption, especially among student users who may be discouraged by overly complex or poorly designed interfaces.
Conversely, slightly lower scores were observed in statements related to confidence in using the system .
and the meaningful organization of data .
This suggests that while the system is functionally adequate, there may be room to enhance the clarity of the interface or provide better onboarding or help features to assist first-time users.
Implications for Student Financial Behavior The system addresses key challenges in student financial management, such as the lack of routine expense tracking and difficulties in budgeting.
By automating data input through OCR and visualizing spending patterns, students can better understand their financial habits.
This is particularly relevant in promoting financial literacy and responsible spending, especially within higher education environments where students often manage finances independently for the first time.
The system's ability to highlight budget overruns and provide visual summaries encourages proactive financial planning, aligning with the goals of educational institutions aiming to cultivate financially responsible graduates.
Limitations and Areas for Improvement Despite the positive results, several limitations were identified:
OCR Accuracy with Poor Image Quality: While the OCR feature performed well under normal conditions, the system may struggle with low-resolution or crumpled receipts.
Enhancing OCR preprocessing .
, contrast adjustment, noise reductio.
could improve accuracy.
Ahmad Fadli Saad.
Muhammad Hairil Shaharudin.
Achmad Yani.
Abdi Manaf.
Development of a Student Expense Tracking System Using Optical Character Recognition.
International Journal of Artificial Intelligence, vol.
12, no.
1, pp.
June 2025.
DOI: 10.
36079/lamintang.
Lack of Real-time Notifications: The system currently lacks features such as alerts for budget overspending or unusual transactions.
Integrating push notifications could enhance real-time financial awareness.
Limited Financial Categorization: The system does not automatically categorize expenses .
food, transport, study material.
, which could improve user insights.
Machine learning or keyword-based categorization could be considered in future development.
The discussion has highlighted the effectiveness of the OCR-based student expense tracking system in terms of both functionality and usability.
The positive feedback from users affirms the value of such a tool in enhancing student financial awareness and promoting better budgeting habits.
Functional robustness, ease of use, and high satisfaction levels all point toward the systemAos potential scalability and integration within university ecosystems.
Future improvements can address identified limitations and further align the system with user needs.
Conclusion The development of the Student Expense Tracking System using OCR has successfully met its three main objectives.
First, the system was effectively designed with clear visual mockups and wireframes that guided its development.
Second, the integration of Optical Character Recognition (OCR) allowed users to extract and convert printed receipt data into digital expense entries, automating what is typically a manual process.
Finally, thorough functional testing was conducted using 10 test scenarios, all of which passed successfully, confirming that core functionalitiesAisuch as login, image upload.
OCR processing, expense input, budget setup, and visualizationAiworked as intended.
During functionality testing, each module produced the expected outcomes.
For instance, the OCR engine was able to extract the total amount accurately in 83.
3% of the tested images, although it faced challenges in cases with poor lighting or distorted fonts.
Usability testing was conducted using the Post-Study System Usability Questionnaire (PSSUQ) among students at Universiti Teknologi MARA (UiTM).
The system scored an overall satisfaction rating of 5.
62 out of 7, indicating a high level of user acceptance.
In particular, users appreciated the systemAos ease of use and efficiency in tracking expenses, but some noted areas for improvement in OCR accuracy and UI responsiveness.
For future development, enhancements should focus on integrating Natural Language Processing (NLP) and Machine Learning (ML).
NLP can improve semantic understanding of extracted text, enabling the system to categorize expenses .
, food, utilities, transportatio.
without user input.
models, trained on labeled receipt data, can boost prediction accuracy and accommodate a wider range of receipt formats.
These technologies will not only increase system intelligence but also help students gain more meaningful insights into their financial habits with minimal effort.
References