Analisis Performa Model TrOCR dalam mengenali Tulisan Tangan pada Catatan Transaksi Manual
DOI:
https://doi.org/10.29407/hq8c4g73Abstract
Pencatatan transaksi secara manual pada usaha kecil seringkali menghadapi kendala dalam kecepatan pencarian data dan akurasi input digital. Penelitian ini bertujuan untuk menganalisis performa model TrOCR (Transformer-based Optical Character Recognition) dalam mendigitalisasi tulisan tangan pada catatan transaksi manual. Metode eksperimen dilakukan melalui proses fine-tuning pada model pretrained microsoft/trocr-base-handwritten dengan dataset sebanyak 666 citra yang telah melalui tahapan preprocessing seperti resize, augmentasi, grayscale, dan adaptive threshold. Evaluasi model dilakukan menggunakan metode 5-Fold Cross Validation dengan metrik Character Error Rate (CER) dan Word Error Rate (WER). Hasil penelitian menunjukkan bahwa model mampu mengenali tulisan tangan dengan sangat baik, di mana performa terbaik dicapai pada Fold 3 dengan nilai CER 2,71%, WER 12,42%, dan akurasi karakter sebesar 97,29%. Temuan ini mengindikasikan bahwa model TrOCR memiliki potensi besar untuk diimplementasikan dalam sistem otomatisasi administrasi transaksi berbasis tulisan tangan.
Keywords:
Adaptive Threshold, Catatan Transaksi, Deep Leraning, TrOCR, Tulisan Tangan.##plugins.themes.default.displayStats.downloads##
References
[1] R. A. Pebriani, T. Yustini, R. Sari, and N. Kholis, "Smart Cooperative: Pelatihan implementasi aplikasi digital untuk meningkatkan efektivitas pengelolaan koperasi," J. Abdimas Ekon. Dan Bisnis, vol. 5, no. 1, pp. 58-65, 2025.
[2] A. B. Setiawan and M. Firmansyah, "Pengembangan Aplikasi Sederhana Pencatatan Keuangan untuk Meningkatkan Efisiensi Administrasi di SMK Al-Khaeriyah Pengampelan," NuCSJo Nusant. Community Serv. J., vol. 2, no. 2, pp. 136-140, 2025.
[3] E. Maulidia, S. Wasiyanti, and K. Miharja, "Perbandingan pencatatan keuangan manual dengan menggunakan zahir accounting pada Warung Kopi Rakjat," Artik. Ilm. Sist. Inf. Akunt., vol. 3, no. 2, pp. 97-102, 2023.
[4] M. R. Firmansyah, A. C. Santoso, A. Farah, U. Monalissa, and M. R. Adiyanto, "Pengaruh pencatatan akuntansi manual dengan pencatatan digital di era globalisasi dalam suatu usaha Snack Rehan Demangan Bangkalan," J. Media Akad., vol. 2, no. 7, 2024.
[5] K. Christianto et al., "RANCANG BANGUN APLIKASI PENCATATAN TRANSAKSI KEUANGAN PADA UMKM KEJUMPA," JATI (Jurnal Mhs. Tek. Inform.), vol. 10, no. 2, pp. 3433-3440, 2026.
[6] L. Abdiansah, S. Sumarno, A. Eviyanti, and N. L. Azizah, "Penerapan Algoritma Convolutional Neural Networks untuk Pengenalan Tulisan Tangan Aksara Jawa," MALCOM Indones. J. Mach. Learn. Comput. Sci., vol. 5, no. 2, pp. 496-504, 2025.
[7] E. Meoded, "Handwritten text recognition of historical manuscripts using transformer-based models," arXiv Prepr. arXiv2508.11499, 2025.
[8] L. Rahmawati, W. T. Atmojo, E. P. Cynthia, M. M. Cynthia, and D. N. Cynthia, "Pembandingan Arsitektur Transformer dan CNN untuk Pengolahan Data Non-Visual," J. Ilmu Komput. dan Tek. Inform., vol. 2, no. 1, pp. 8-14, 2026.
[9] F. Safira, "Studi Komparatif Model OCR Berbasis AI untuk Dokumen Cetak dan Tulisan Tangan," Int. J. Informatics, vol. 1, no. 1, 2025.
[10] S. B. Bhaskoro, R. A. Pratama, and S. A. H. Darmawan, "Deteksi dan Interpretasi Tulisan Tangan Bahasa Indonesia melalui Pemrosesan Citra dan Optical Character Recognition (OCR)," JTRM (Jurnal Teknol. dan Rekayasa Manufaktur), vol. 7, no. 1, pp. 48-64, 2025.
[11] M. R. Anggraeni and B. S. W. Poetro, "Penerapan Deep Convolutional Autoencoder dengan Skip Connection Untuk Menghilangkan Derau pada Citra Digital," Bridg. J. Publ. Sist. Inf. dan Telekomun., vol. 4, no. 1, pp. 10-28, 2026.
[12] H. Zhang, E. Whittaker, and I. Kitagishi, "Extending TrOCR for text localization-free OCR of full-page scanned receipt images," in Proceedings of the IEEE/CVF International Conference on Computer Vision, 2023, pp. 1479-1485.
[13] D. Anggreani and L. Lukman, "Peningkatan metode YOLOv7 dengan proses augmentasi image pada klasifikasi jenis kupu-kupu," J. Teknol. Sist. Inf., vol. 4, no. 2, pp. 243-253, 2023.
[14] R. A. Tavares, "Comparison of image preprocessing techniques for vehicle license plate recognition using ocr: Performance and accuracy evaluation," arXiv Prepr. arXiv2410.13622, 2024.
[15] D. Situmorang, J. Lumban-Gaol, and E. Tambunan, "DETEKSI TUMPAHAN MINYAK MENGGUNAKAN METODE ADAPTIVE THRESHOLD PADA CITRA SATELIT SENTINEL-1A DI PERAIRAN UTARA KARAWANG," J. Rekayasa Tek. Sipil dan Lingkungan-CENTECH, vol. 6, no. 1, pp. 28-36, 2025.
[16] W. Wijiyanto, A. I. Pradana, S. Sopingi, and V. Atina, "Teknik K-Fold Cross Validation untuk Mengevaluasi Kinerja Mahasiswa," J. Algoritm., vol. 21, no. 1, pp. 239-248, 2024.
[17] S. Ouzerrout, "OCER and OCWER: Integrating Visual Similarity and Segmentation in OCR Evaluation," in Second Workshop on Language Models for Underserved Communities (LM4UC), 2026.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Authors who publish with this journal agree to the following terms:
- Copyright on any article is retained by the author(s).
- The author grants the journal, right of first publication with the work simultaneously licensed under a Creative Commons Attribution License that allows others to share the work with an acknowledgment of the work’s authorship and initial publication in this journal.
- Authors are able to enter into separate, additional contractual arrangements for the non-exclusive distribution of the journal’s published version of the work (e.g., post it to an institutional repository or publish it in a book), with an acknowledgment of its initial publication in this journal.
- Authors are permitted and encouraged to post their work online (e.g., in institutional repositories or on their website) prior to and during the submission process, as it can lead to productive exchanges, as well as earlier and greater citation of published work.
- The article and any associated published material is distributed under the Creative Commons Attribution-ShareAlike 4.0 International License
