EKSTRAKSI INFORMASI STRUK BELANJA MELALUI PEMANFAATAN TESSERACT DAN REGULAR EXPRESSION MENJADI DATA TERSTRUKTUR

Amelia Ananda Setiawan, - and Rangga Gelar Guntara, - and Btari Mariska Purwaamijaya, - (2025) EKSTRAKSI INFORMASI STRUK BELANJA MELALUI PEMANFAATAN TESSERACT DAN REGULAR EXPRESSION MENJADI DATA TERSTRUKTUR. S1 thesis, Universitas Pendidikan Indonesia.

Abstract

Belanja merupakan aktivitas pengeluaran rutin yang menghasilkan catatan transaksi dalam bentuk struk belanja. Struk tersebut memuat berbagai informasi seperti nama produk, kuantitas, harga satuan, total, dan diskon, yang tidak hanya sebagai bukti transaksi, tetapi memiliki manfaat lain untuk memantau pengeluaran melalui pencatatan keuangan. Namun, dalam praktiknya, pencatatan keuangan secara manual melalui kertas dianggap menyulitkan karena rentan hilang. Pencatatan menggunakan aplikasi digital pun tetap menjadi hambatan karena pengguna perlu memasukan data secara manual. Penelitian ini bertujuan untuk membangun sistem ekstraksi struk belanja melalui penerapan Tesseract, YOLO, dan Regular Expression. Sistem dikembangkan menggunakan metode Software Development Life Cycle. Hasil evaluasi model YOLO menunjukkan performa yang cukup baik dengan f1-score 0.88, mAp@50 0.93, dan mAp@50-95 0.67. Hasil pengujian alpha menggunakan blackbox testing menunjukkan bahwa sistem berjalan sesuai fungsionalitasnya. Sementara itu, hasil pengujian beta memperoleh skor rata-rata 4,47 dari skala 5 yang menunjukkan sistem dinilai mampu membantu pengguna dalam pencatatan keuangan karena lebih cepat, mengurangi beban input manual, dan mudah digunakan. Shopping is a routine expenditure activity that generates transaction records in the form of receipts. These receipts contain various information such as product names, quantities, unit prices, totals, and discounts, which not only serve as proof of transactions but also have additional benefits for monitoring expenses through financial recordkeeping. However, in practice, manual financial recordkeeping using paper is considered inconvenient due to its susceptibility to loss. Even when using digital applications, it still poses challenges as users must enter the data manually. This study aims to develop a receipt extraction system by applying Tesseract, YOLO, and Regular Expressions. The system was developed using the Software Development Life Cycle method. The evaluation results of the YOLO model demonstrated good performance, with an F1-score of 0.88, mAP@50 of 0.93, and mAP@50–95 of 0.67. Alpha testing using blackbox testing showed that the system operated according to its intended functionalities. Meanwhile, beta testing achieved an average score of 4.47 out of 5, indicating that the system is perceived as effective in assisting users with financial recordkeeping by enabling faster processing, reducing manual input workload, and providing ease of use.

	Text S_BIDI_2104440_Title.pdf Download (1MB)
	Text S_BIDI_2104440_Chapter1.pdf Download (752kB)
	Text S_BIDI_2104440_Chapter2.pdf Restricted to Staf Perpustakaan Download (1MB)
	Text S_BIDI_2104440_Chapter3.pdf Download (821kB)
	Text S_BIDI_2104440_Chapter4.pdf Restricted to Staf Perpustakaan Download (1MB)
	Text S_BIDI_2104440_Chapter5.pdf Download (474kB)
	Text S_BIDI_2104440_Appendix.pdf Restricted to Staf Perpustakaan Download (1MB)

Item Type:	Thesis (S1)
Additional Information:	https://scholar.google.com/citations?view_op=list_works&hl=en&user=FZmdc34AAAAJ ID SINTA Dosen Pembimbing: Rangga Gelar Guntara: 6738149 Btari Mariska Purwaamijaya: 6681170
Uncontrolled Keywords:	Struk Belanja, Pencatatan Keuangan, Tesseract, YOLO, Regular Expression Purchase Receipt, Financial Recording, Tesseract, YOLO, Regular Expression
Subjects:	L Education > L Education (General)
Divisions:	UPI Kampus Tasikmalaya > S1 Bisnis Digital
Depositing User:	Amelia Ananda Setiawan
Date Deposited:	16 Sep 2025 06:19
Last Modified:	16 Sep 2025 06:19
URI:	http://repository.upi.edu/id/eprint/139144

Actions (login required)

View Item