Receipt Scanning with EasyOCR and ChatGPT-4o in a Mobile Finance App: an Agile Kanban Approach
Year: 2025
Type: article
Abstract: Technological advancements have provided convenience for Generation Z in managing finances; however, many are still not accustomed to recording their financial activities regularly. Shopping receipts, which should serve as proof of transactions, are often ignored or poorly managed, despite their important role in tracking expenses. Therefore, this research aims to develop an Android-based financial recording application capable of handling both manual input and automated recording through receipt scanning using Optical Character Recognition (OCR) technology. The findings indicate that ChatGPT-4o significantly outperforms EasyOCR by providing more consistent accuracy and faster, stable processing, making it a more reliable solution for receipt-based financial recording. Developed using the Agile Kanban method, the application was validated through alpha testing and proven to function properly across all features. Beyond practical benefits for users, this research also contributes to the financial technology literature by demonstrating the integration of large language models (LLM) to enhance OCR performance in mobile finance applications.
Source: Jurnal Teknologi Terapan G-Tech
Institution: Universitas Muslim Indonesia
Language: English
Subfield: Computer Vision and Pattern Recognition
Field: Computer Science
Domain: Physical Sciences
Open Access status: gold
Estimated APC paid: $19
OpenAlex ID: W4414807168
FWCI:
Cites:
Cited by:
Related to: