| Article name |
Automatic Bank Slip Data Extraction System via Tesseract OCR and QR Code Decoding
|
|
|
|
| Article type |
Research article
|
| Authors |
Sorawit Bunset(1), Tipyada Kaewmakam(1), Part Pramokchon(1),
Attawit Changkamanon(1) and Kongkarn Dullayachai(1*)
|
| Office |
Computer Science Department, Faculty of Science, Maejo University(1) *Corresponding author: kongkarn@gmaejo.mju.ac.th
|
| Journal name |
Vol. 12 No.2 (2026): May - August
|
| Abstract |
This research aims to develop and compare the efficiency of an automated bank transfer receipt data extraction system using Tesseract OCR technology combined with QR Code decoding. The system was developed using Go 1.22 and integrated with a LINE Official Account via Webhook to provide near real-time services and store data in a PostgreSQL database.
The evaluation was conducted using 102 sample receipts from three banks, with 34 samples from each. The transaction reference numbers extracted by OCR were compared with the data decoded from the QR Code using the Windowed Levenshtein Distance technique, with a similarity threshold of 70 percent. The results, categorized by bank, showed that Kasikornbank achieved 100% accuracy with an average processing time of 1.65 seconds; Krungthai Bank achieved 97.06% accuracy with an average processing time of 1.78 seconds; and Siam Commercial Bank achieved 100% accuracy with an average processing time of 1.84 seconds. In summary, the system achieved an overall average accuracy of 99.02% and a total average processing time of 1.76 seconds (S.D. 0.50 seconds) per receipt. These findings demonstrate the system's high efficiency, making it suitable for application in automated payment verification workflows for organizations.
|
| Keywords |
Bank Transfer Receipt; Tesseract OCR; QR Code; Windowed Levenshtein Distance; LINE Official Account
|
| Page number |
81-100
|
| ISSN |
ISSN 3027-7280 (Online)
|
| DOI |
|
| ORCID_ID |
0009-0005-2802-2871
|
| Article file |
https://mitij.mju.ac.th/ARTICLE/R69054.pdf
|
| | |
| Reference | |
| |
ณรงค์เกียรติ นามห้วยทอง และคณะ. (2568). การศึกษาประสิทธิภาพของ Tesseract OCRสำหรับการประมวลผลภาพในการตรวจสอบธุรกรรมทางการเงิน. วารสารแม่โจ้เทคโนโลยีสารสนเทศและนวัตกรรม, 9(2), 51–60.
https://mitij.mju.ac.th/Search_Detail_Journal_MJU.aspx?Herb_ID=0209B
|
| |
วิศรุต เหล่าดารา. (2565). ระบบเซ็นเซอร์ชื่อบุคคลออกจากเอกสารสแกนคำพิพากษาด้วย ปัญญาประดิษฐ์ (สารนิพนธ์ปริญญามหาบัณฑิต). มหาวิทยาลัยธุรกิจบัณฑิตย์ https://libdoc.dpu.ac.th/thesis/Wisarut.Kang.pdf
|
| |
Google. (n.d.). Tesseract OCR.
Retrieved from https://github.com/tesseract-ocr/tesseract
|
| |
Levenshtein, V. I. (1966). Binary codes capable of correcting deletions, insertions, and reversals. Soviet Physics Doklady, 10(8), 707–710. https://nymity.ch/sybilhunting/pdf/Levenshtein1966a.pdf
|
| |
Sayallar, C., Sayar, A., & Babalik, N. (2023). An OCR engine for printed receiptImages using deep learning techniques. International Journal of Advanced Computer Science and Applications (IJACSA), 14(2).
https://thesai.org/Publications/ViewPaper?Volume=14&Issue=2&Code=IJACSA&SerialNo=95
|
| |
Thammarak, K., Kongkla, P., Sirisathitkul, Y., & Intakosum, S. (2022). Comparative analysis of Tesseract and Google Cloud Vision for Thai vehicleregistration certificate. International Journal of Electrical and Computer Engineering (IJECE), 12(2), 1849-1858. http://doi.org/10.11591/ijece.v12i2.pp1849-1858
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|
| |
|