Pengenalan Karakter Angka pada Meter-Air Analog Menggunakan Arsitektur Convolutional Neural Network (CNN)
Keywords:
meter air analog, deteksi objek, CNN, Yolov10Abstract
Sebagian besar Perusahaan Air Minum (PAM) di Indonesia hingga kini masih mengandalkan meter air analog bertipe flow meter turbin, bahkan pada wilayah operasional berskala besar dengan jumlah pelanggan aktif yang melebihi 100.000. Kendala utama dalam transisi ke teknologi yang lebih modern, seperti meter air pintar, adalah tingginya biaya pengadaan dan implementasinya. Akibatnya, proses pencatatan angka pemakaian air masih dilakukan secara manual setiap bulan, yang memerlukan keterlibatan tenaga kerja dalam jumlah besar dan tidak memungkinkan verifikasi pembacaan secara masif. Hal ini seringkali berdampak pada rendahnya akurasi data pencatatan, serta menimbulkan ketidakpuasan pelanggan akibat tagihan yang tidak sesuai dengan konsumsi riil.
Untuk menjawab tantangan tersebut, kami mengusulkan solusi berbasis visi komputer dengan menerapkan model deteksi objek You Only Look Once versi 10 (YOLOv10), sebuah arsitektur berbasis Convolutional Neural Network (CNN) yang dirancang untuk kebutuhan deteksi waktu nyata (real-time). YOLOv10 mengusung pendekatan Non- Maximum Suppression (NMS)-free, yang memungkinkan kecepatan inferensi lebih tinggi dibandingkan generasi sebelumnya, tanpa mengorbankan akurasi deteksi. Berdasarkan hasil pengujian pada citra meter air analog, model YOLOv10 menunjukkan performa yang menjanjikan meskipun hanya dijalankan pada CPU. Model ini mampu mencapai nilai precision sebesar 0,9426 (94,26%) dan recall sebesar 0,9783 (97,83%) dengan rata-rata waktu inferensi 1,7157 detik per gambar. Temuan ini menunjukkan potensi penerapan teknologi deteksi otomatis dalam mendukung efisiensi operasional PAM dan meningkatkan akurasi pembacaan meter air secara signifikan.
References
[1] N. Saetan and K. Dittakan, “EasyChair Preprint Thailand Water Meter Reading Using Convolutional Neural Networks from Smartphone Imagery Thailand Water Meter Reading Using Convolutional Neural Networks From Smartphone Imagery,” 2024.
[2] Y. Liang, Y. Liao, S. Li, W. Wu, T. Qiu, and W. Zhang, “Research on water meter reading recognition based on deep learning,” Sci Rep, vol. 12, no. 1, Dec. 2022, doi: 10.1038/s41598-022-17255-3.
[3] Y. Wang and X. Xiang, “GMS-YOLO Improved water meter reading recognition algorithm based on YOLOv8,” 2024, doi: 10.21203/rs.3.rs-4453194/v1.
[4] L. Sun, Y. Yuan, S. Qiao, and R. Qi, “Detection of Water Meter Digits Based on Improved Faster R-CNN,” Journal of Computer and Communications, vol. 12, no. 03, pp. 1–13, 2024, doi: 10.4236/jcc.2024.123001.
[5] A. Wang et al., “YOLOv10: Real-Time End-to-End Object Detection,” May 2024, [Online]. Available: http://arxiv.org/abs/2405.14458
[6] T. Yang, S. Zhou, A. Xu, J. Ye, and J. Yin, “An Approach for Plant Leaf Image Segmentation Based on YOLOV8 and the Improved DEEPLABV3+,” Plants, vol. 12, no. 19, Oct. 2023, doi: 10.3390/plants12193438.
[7] J. Terven and D. Cordova-Esparza, “A Comprehensive Review of YOLO Architectures in Computer Vision: From YOLOv1 to YOLOv8 and YOLO-NAS,” Apr. 2023, doi: 10.3390/make5040083.
[8] O. M. Khare, S. Gandhi, A. M. Rahalkar, and S. Mane, “YOLOv8-Based Visual Detection of Road Hazards: Potholes,
Sewer Covers, and Manholes,” Oct. 2023, [Online]. Available: http://arxiv.org/abs/2311.00073
[9] D. Reis, J. Kupec, J. Hong, and A. Daoudi, “Real-Time Flying Object Detection with YOLOv8,” May 2023, [Online].
Available: http://arxiv.org/abs/2305.09972
[10] Z. Huang, L. Li, G. C. Krizek, and L. Sun, “Research on Traffic Sign Detection Based on Improved YOLOv8,” Journal of Computer and Communications, vol. 11, no. 07, pp. 226–232, 2023, doi: 10.4236/jcc.2023.117014.
[11] R. Jia, B. Lv, J. Chen, H. Liu, L. Cao, and M. Liu, “Underwater Object Detection in Marine Ranching Based on Improved YOLOv8,” J Mar Sci Eng, vol. 12, no. 1, Jan. 2024, doi: 10.3390/jmse12010055.
[12] Q. Liu, Y. Liu, and D. Lin, “Revolutionizing Target Detection in Intelligent Traffic Systems: YOLOv8-SnakeVision,”
Electronics (Switzerland), vol. 12, no. 24, Dec. 2023, doi: 10.3390/electronics12244970.
[13] J. Moosmann et al., “Ultra-Ef cient On-Device Object Detection on AI-Integrated Smart Glasses with TinyissimoYOLO,” 2023.
[14] X. Wang, H. Gao, Z. Jia, and Z. Li, “BL-YOLOv8: An Improved Road Defect Detection Model Based on YOLOv8,”
Sensors (Basel), vol. 23, no. 20, Oct. 2023, doi: 10.3390/s23208361.
[15] A. Dumitriu, F. Tatui, F. Miron, R. T. Ionescu, and R. Timofte, “Rip Current Segmentation: A Novel Benchmark and YOLOv8 Baseline Results,” 2023. [Online]. Available: www.noaa.gov/.
[16] Y. Chen, Q. Chen, Q. Hu, and J. Cheng, “DATE: Dual Assignment for End-to-End Fully Convolutional Object Detection,” Nov. 2022, [Online]. Available: http://arxiv.org/abs/2211.13859
[17] F. Wang, Q. Bao, Z. Wang, and Y. Chen, “Optimizing Transformer based on high-performance optimizer for predicting employment sentiment in American social media content.”
[18] A. Wang et al., “Mitigating the Negative Impact of Over-association for Conversational Query Production,” Sep. 2024, [Online]. Available: http://arxiv.org/abs/2409.19572
[19] J. Redmon and A. Farhadi, “YOLOv3: An Incremental Improvement,” Apr. 2018, [Online]. Available: http://arxiv.org/abs/1804.02767
Downloads
Published
How to Cite
Issue
Section
License
This work is licensed under a Creative Commons Attribution-ShareAlike 4.0 International License.
