Deteksi Informasi Kadar Natrium pada Label Produk Kemasan Menggunakan FPN, Faster R-CNN, dan OCR
Keywords:
sodium, Feature Pyramid Network (FPN), Convolutional Neural Network (CNN), Optical Character Recognition (OCR), nutrition facts, Natrium, informasi giziAbstract
Informasi nilai gizi, khususnya kandungan natrium, berperan penting dalam mendorong masyarakat memilih makanan yang lebih sehat. Namun, kesadaran terhadap label gizi masih rendah sehingga sering diabaikan. Sistem otomatis dikembangkan untuk mendeteksi dan mengekstraksi informasi kadar natrium dari citra kemasan produk. Sistem menggunakan pendekatan deteksi objek berbasis Feature Pyramid Network (FPN) dan Faster R-CNN melalui framework Detectron2. Ekstraksi teks kadar natrium dilakukan menggunakan Tesseract OCR. Sebanyak 3.028 gambar digunakan dalam proses pelatihan, validasi, dan pengujian FPN. Pada tahap pelatihan diperoleh akurasi deteksi 98,90% (AP50) dan (AR) 85,00%. Sementara itu, pengujian OCR terhadap 400 gambar berbeda dari data pelatihan FPN, menghasilkan rata-rata akurasi ekstraksi teks 85,13%. Kesalahan pengenalan umumnya disebabkan kualitas citra rendah, seperti gambar blur, orientasi label miring, pantulan cahaya, atau teks terlalu tebal. Meskipun terdapat keterbatasan, sistem menunjukkan potensi kuat sebagai alat otomatis membaca informasi gizi, khususnya kandungan natrium, dengan catatan diperlukan penyempurnaan lebih lanjut agar konsisten di kondisi nyata.
References
L. Hasibuan, “Batas Aman Konsumsi Gula dan Garam Harian Menurut WHO,” CNBC Indonesia. Diakses: 3 Maret 2025. [Daring]. Tersedia pada: https://www.cnbcindonesia.com/lifestyle/20241203142608-33-593000/batas-aman-konsumsi-gula-dan-garam-harian-menurut-who
D. Arlinta, “Lebih dari 50 Persen Masyarakat Indonesia Konsumsi Garam Berlebihan,” kompas.id. Diakses: 3 Maret 2025. [Daring]. Tersedia pada: https://www.kompas.id/baca/humaniora/2023/03/28/lebih-dari-50-persen-masyarakat-indonesia-konsumsi-garam-berlebihan
X. Shao dkk., “Multi-Scale Feature Pyramid Network: A Heavily Occluded Pedestrian Detection Network Based on ResNet,” Sensors, vol. 21, no. 5, hlm. 1820, Jan 2021, doi: 10.3390/s21051820.
W. Sheng, X. Yu, J. Lin, dan X. Chen, “Faster RCNN Target Detection Algorithm Integrating CBAM and FPN,” Applied Sciences, vol. 13, no. 12, hlm. 6913, Jan 2023, doi: 10.3390/app13126913.
J. Kang, M. Ibrayim, dan A. Hamdulla, “MR-FPN: Multi-Level Residual Feature Pyramid Text Detection Network Based on Self-Attention Environment,” Sensors (Basel), vol. 22, no. 9, hlm. 3337, Apr 2022, doi: 10.3390/s22093337.
J. Pardede dan H. Hardiansah, “Deteksi Objek Kereta Api menggunakan Metode Faster R-CNN dengan Arsitektur VGG 16,” MIND (Multimedia Artificial Intelligent Networking Database) Journal, vol. 7, no. 1, hlm. 21–36, Jun 2022, doi: 10.26760/mindjournal.v7i1.21-36.
J. Wang, J. Tang, M. Yang, X. Bai, dan J. Luo, “Improving OCR-based Image Captioning by Incorporating Geometrical Relationship,” dalam 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), Nashville, TN, USA: IEEE, Jun 2021, hlm. 1306–1315. doi: 10.1109/CVPR46437.2021.00136.
T. M. E. Saputra, A. Susanto, dan B. J. Carmelita, “Implementation of Tesseract OCR and Bounding Box for Text Extraction on Food Nutrition Labels,” Building of Informatics, Technology and Science (BITS), vol. 6, no. 3, hlm. 1403–1412, Des 2024, doi: 10.47065/bits.v6i3.6107.
K. A. Nugraha, “Penerapan Optical Character Recognition untuk Pengenalan Variasi Teks pada Media Presentasi Pembelajaran,” Jurnal Buana Informatika, vol. 15, no. 01, hlm. 69–78, Apr 2024, doi: https://orcid.org/0000-0003-3401-8033.
Fino Charli, Hadi Syaputra, Muhammad Akbar, Siti Sauda, dan Febriyanti Panjaitan, “Implementasi Metode Faster Region Convolutional Neural Network (Faster R-CNN) Untuk Pengenalan Jenis Burung Lovebird,” vol. 1, no. 3, Des 2020, doi: 10.51519/journalita.volume1.isssue3.year2020.page185-197.
B. Yu, X. Pan, X. Li, G. Liu, dan Y. Ma, “Object Detection Algorithm Based on Improved Feature Pyramid,” Scientific Programming, vol. 2022, no. 1, hlm. 3583399, 2022, doi: 10.1155/2022/3583399.
R. D. Altarez, “Faster R–CNN, RetinaNet and Single Shot Detector in different ResNet backbones for marine vessel detection using cross polarization C-band SAR imagery,” Remote Sensing Applications: Society and Environment, vol. 36, hlm. 101297, Nov 2024, doi: 10.1016/j.rsase.2024.101297.
S. Hartanto, A. Sugiharto, dan S. N. Endah, “Optical Character Recognition Menggunakan Algoritma Template Matching Correlation,” Jurnal Masyarakat Informatika, vol. 5, no. 9, hlm. 1–12, Apr 2015, doi: 10.14710/jmasif.5.9.8435.
F. A. Rafi, A. Fanggidae, dan Y. T. Polly, “Sistem Pendeteksi Kerusakan Jalan Aspal Menggunakan Canny Edge Detection,” J-Icon : Jurnal Komputer dan Informatika, vol. 11, no. 1, hlm. 85–90, Mar 2023, doi: 10.35508/jicon.v11i1.10100.
A. M. Dwitala, A. Fanggidae, Y. T. Polly, dan B. B. Koten, “Klasifikasi Citra Biji Arbila di Kabupaten Kupang menggunakan k-Nearest Neighbor berdasarkan fitur Warna dan Tekstur,” Jurnal Ilmiah Telekomunikasi, Kendali dan Elektronika Terapan, vol. 12, no. 2, hlm. 177–186, Okt 2024, doi: 10.34010/telekontran.v12i2.14207.
A. Ridhovan dan A. Suharso, “Penerapan Metode Residual Network (Resnet) dalam Klasifikasi Penyakit pada Daun Gandum,” JIPI (Jurnal Ilmiah Penelitian dan Pembelajaran Informatika), vol. 7, no. 1, hlm. 58–65, Feb 2022, doi: 10.29100/jipi.v7i1.2410.
G. Zhao, W. Ge, dan Y. Yu, “GraphFPN: Graph Feature Pyramid Network for Object Detection,” 8 Januari 2022, arXiv:2108.00580. doi: 10.48550/arXiv.2108.00580.
G. Yang, J. Lei, Z. Zhu, S. Cheng, Z. Feng, dan R. Liang, “AFPN: Asymptotic Feature Pyramid Network for Object Detection,” 24 September 2023. doi: 10.48550/arXiv.2306.15988.
X. Dong, Y. Qin, Y. Gao, R. Fu, S. Liu, dan Y. Ye, “Attention-Based Multi-Level Feature Fusion for Object Detection in Remote Sensing Images,” Remote Sensing, vol. 14, no. 15, hlm. 3735, Jan 2022, doi: 10.3390/rs14153735.
S.-H. Tsang, “Review: Faster R-CNN (Object Detection),” TDS Archive. Diakses: 20 Juli 2025. [Daring]. Tersedia pada: https://medium.com/data-science/review-faster-r-cnn-object-detection-f5685cb30202
T. Bai dkk., “An Optimized Faster R-CNN Method Based on DRNet and RoI Align for Building Detection in Remote Sensing Images,” Remote Sensing, vol. 12, no. 5, hlm. 762, Jan 2020, doi: 10.3390/rs12050762.
I. H. A. Amin dan A. Aprilino, “Implementasi Algoritma Yolo dan Tesseract OCR pada Sistem Deteksi Plat Nomor Otomatis,” Jurnal Teknoinfo, vol. 16, no. 1, hlm. 54–59, Jan 2022, doi: 10.33365/jti.v16i1.1522.
Downloads
Published
Issue
Section
License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
Copyright of this journal is assigned to Jurnal Buana Informatika as the journal publisher by the knowledge of author, whilst the moral right of the publication belongs to author. Every printed and electronic publications are open access for educational purposes, research, and library. The editorial board is not responsible for copyright violation to the other than them aims mentioned before. The reproduction of any part of this journal (printed or online) will be allowed only with a written permission from Jurnal Buana Informatika.
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
