Optimasi MobileNetV2 untuk Klasifikasi Hama dan Simulasi Dampak Finansial
DOI:
https://doi.org/10.35889/jutisi.v15i4.3897Keywords:
MobileNetV2, klasifikasi citra, hama pertanian, data augmentation, dampak finansialAbstract
Inaccurate pest identification can delay control measures and increase the risk of agricultural losses. This study evaluated MobileNetV2 for image-based pest classification and simulated the financial impact of prediction errors. The Dangerous Farm Insects Dataset, containing 1,591 images in 15 classes, was used in four scenarios: baseline, augmentation, baseline fine-tuning, and augmentation fine-tuning. The models were assessed using accuracy, precision, recall, macro F1-score, and confusion matrices. The results showed that the baseline achieved the best performance, with an accuracy of 0.750 and a macro F1-score of 0.744. Fine-tuning produced the same accuracy but a slightly lower macro F1-score, while the two augmentation scenarios achieved accuracies of only 0.528 and 0.536. Most errors involved visually similar classes. The financial simulation indicated that the baseline had the lowest relative risk, although the cost values remained illustrative parameters requiring field validation.
Keywords: MobileNetV2; image classification; agricultural pests; data augmentation; financial impact
Abstrak
Identifikasi hama yang tidak tepat dapat menunda pengendalian dan meningkatkan risiko kerugian pertanian. Penelitian ini mengevaluasi MobileNetV2 untuk klasifikasi citra hama serta menyimulasikan dampak finansial kesalahan prediksi. Dangerous Farm Insects Dataset yang berisi 1.591 citra dalam 15 kelas digunakan pada empat skenario: baseline, augmentation, baseline fine-tuning, dan augmentation fine-tuning. Model dinilai menggunakan accuracy, precision, recall, macro F1-score, dan confusion matrix. Hasil pengujian menunjukkan bahwa baseline memberikan performa terbaik dengan accuracy 0,750 dan macro F1-score 0,744. Fine-tuning menghasilkan accuracy yang sama, tetapi macro F1-score sedikit lebih rendah, sedangkan dua skenario augmentation hanya mencapai accuracy 0,528 dan 0,536. Kesalahan terbanyak terjadi pada kelas yang serupa secara visual. Simulasi finansial menunjukkan bahwa baseline memiliki risiko relatif terendah, meskipun nilai biaya masih berupa parameter ilustratif dan memerlukan validasi lapangan.
Kata kunci: MobileNetV2; klasifikasi citra; hama pertanian; data augmentation; dampak finansial
References
[1] S. Savary et al., "The global burden of pathogens and pests on major food crops," Nat. Ecol. Evol., vol. 3, pp. 430-439, 2019, doi: 10.1038/s41559-018-0793-y.
[2] D. Popescu, A. Dinca, L. Ichim, and N. Angelescu, "New trends in detection of harmful insects and pests in modern agriculture using artificial neural networks: A review," Front. Plant Sci., vol. 14, p. 1268167, 2023, doi: 10.3389/fpls.2023.1268167.
[3] S. M. Venkateswara and J. Padmanabhan, "Deep learning based agricultural pest monitoring and classification," Sci. Rep., vol. 15, p. 8684, 2025, doi: 10.1038/s41598-025-92659-5.
[4] T. Dalal, "Dangerous Farm Insects Dataset," Kaggle, 2024. [Online]. Tersedia: https://www.kaggle.com/datasets/tarundalal/dangerous-insects-dataset
[5] A. Kamilaris and F. X. Prenafeta-Boldu, "Deep learning in agriculture: A survey," Comput. Electron. Agric., vol. 147, pp. 70-90, 2018, doi: 10.1016/j.compag.2018.02.016.
[6] S. Zhao, J. Liu, Z. Bai, C. Hu, and Y. Jin, "Crop pest recognition in real agricultural environment using convolutional neural networks by a parallel attention mechanism," Front. Plant Sci., vol. 13, p. 839572, 2022, doi: 10.3389/fpls.2022.839572.
[7] K. Thenmozhi and U. S. Reddy, "Crop pest classification based on deep convolutional neural network and transfer learning," Comput. Electron. Agric., vol. 164, p. 104906, 2019, doi: 10.1016/j.compag.2019.104906.
[8] D. Xia, P. Chen, B. Wang, J. Zhang, and C. Xie, "Insect detection and classification based on an improved convolutional neural network," Sensors, vol. 18, no. 12, p. 4169, 2018, doi: 10.3390/s18124169.
[9] C. Li, T. Zhen, and Z. Li, "Image classification of pests with residual neural network based on transfer learning," Appl. Sci., vol. 12, no. 9, p. 4356, 2022, doi: 10.3390/app12094356.
[10] M. Sandler, A. Howard, M. Zhu, A. Zhmoginov, and L.-C. Chen, "MobileNetV2: Inverted Residuals and Linear Bottlenecks," in Proc. IEEE Conf. Comput. Vis. Pattern Recognit. (CVPR), 2018, pp. 4510-4520, doi: 10.1109/CVPR.2018.00474.
[11] A. G. Howard et al., "MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications," arXiv preprint arXiv:1704.04861, 2017. [Online]. Tersedia: https://arxiv.org/abs/1704.04861
[12] C. Shorten and T. M. Khoshgoftaar, "A survey on image data augmentation for deep learning," J. Big Data, vol. 6, no. 1, p. 60, 2019, doi: 10.1186/s40537-019-0197-0.
[13] F. Zhuang et al., "A comprehensive survey on transfer learning," Proc. IEEE, vol. 109, no. 1, pp. 43-76, 2021, doi: 10.1109/JPROC.2020.3004555.
[14] M. Grandini, E. Bagli, and G. Visani, "Metrics for Multi-Class Classification: An Overview," arXiv preprint arXiv:2008.05756, 2020. [Online]. Tersedia: https://arxiv.org/abs/2008.05756
[15] Badan Pusat Statistik, "Hasil Survei Ekonomi Pertanian (SEP) 2024: Nilai dan Biaya Produksi Beberapa Komoditas Utama pada Subsektor Tanaman Pangan, Hortikultura, dan Perkebunan," Jakarta, Publikasi No. 05100.25001, 2024. [Online]. Tersedia: https://www.bps.go.id/id/publication/2025/01/24/8faa8436999bb9b54ba0b7a9/hasil-survei-ekonomi-pertanian-sep-2024-nilai-dan-biaya-produksi-beberapa-komoditas-utama-pada-subsektor-tanaman-pangan-hortikultura-dan-perkebunan.html
[16] B. Wang, "Identification of crop diseases and insect pests based on deep learning," Sci. Program., vol. 2022, p. 9179998, 2022, doi: 10.1155/2022/9179998.
[17] X. Wei, J. Wu, and H. Su, "Fine-Grained Image Analysis with Deep Learning: A Survey," IEEE Trans. Pattern Anal. Mach. Intell., vol. 44, no. 12, pp. 8927-8948, 2022, doi: 10.1109/TPAMI.2021.3126648.
[18] T. Kasinathan, D. Singaraju, and S. R. Uyyala, "Insect classification and detection in field crops using modern machine learning techniques," Inf. Process. Agric., vol. 8, no. 3, pp. 446-457, 2021, doi: 10.1016/j.inpa.2020.09.006.
[19] R. Sharma, A. Kamble, and R. Raman, "Deep Learning Applications for Agriculture: A Review," Sensors, vol. 22, no. 18, p. 7004, 2022, doi: 10.3390/s22187004.
[20] S. Utomo, I. M. I. Subroto, and A. Riansyah, "Deteksi Plagiat Tugas Akhir dengan Metode Jaccard Similarity," Jurnal Transistor Elektro dan Informatika (TRANSISTOR EI), vol. 4, no. 2, pp. 132-141, 2022.
[21] S. Utomo, M. N. W. Affandi, and G. F. Shidik, "Asphalt Road Quality Classification Based on CNN Architecture VGG16," in 2023 International Seminar on Application for Technology of Information and Communication (iSemantic), Semarang, Indonesia, 2023.
