Comparative Analysis of YOLOv8 and MobileNetV2 for Digital Image-Based Bird's Eye Chili Leaf Disease Classification

I Gede Tiar Eka Saputra; I Made Gede Sunarya

doi:10.59261/jbt.v7i3.723

Authors

I Gede Tiar Eka Saputra Universitas Pendidikan Ganesha
I Made Gede Sunarya Universitas Pendidikan Ganesha

DOI:

https://doi.org/10.59261/jbt.v7i3.723

Keywords:

5-Fold Cross Validation, YOLOv8-cls, MobileNetV2, Deep Learning, Leaf Disease Classification, Bird's Eye Chili

Abstract

Background: Bird’s eye chili is an important horticultural commodity in Indonesia but is highly vulnerable to leaf diseases such as mosaic virus, leaf spot, and chlorosis. Visual disease identification is often inaccurate and can delay treatment, while previous deep learning studies have mainly used laboratory-acquired datasets with limited applicability to field conditions.

Objective: This study compares the performance of YOLOv8 classification (YOLOv8-cls) and MobileNetV2 in classifying bird’s eye chili leaf diseases using digital images acquired under field conditions.

Methods: A dataset of 800 leaf images representing four classes (healthy, mosaic virus, leaf spot, and chlorosis) was collected under natural field conditions. The images were preprocessed and used to train both models through transfer learning using ImageNet-pretrained weights. Performance was evaluated using stratified 5-fold cross-validation and measured using accuracy, precision, recall, F1-score, and the Matthews correlation coefficient (MCC).

Results: Both models achieved average accuracies above 98%. YOLOv8-cls achieved the best overall performance, with 98.38% accuracy, 98.55% precision, 98.35% recall, and 98.41% F1-score, while requiring only 2.84 MB of storage, making it suitable for deployment on edge devices. MobileNetV2 achieved 98.25% accuracy, 98.36% precision, 98.16% recall, and 98.24% F1-score, while reducing training time by approximately 40%.

Conclusion: YOLOv8-cls is the more suitable architecture for mobile and edge-based plant disease detection because of its superior classification performance and compact model size, whereas MobileNetV2 is advantageous when rapid model retraining is required. These findings provide a practical basis for developing accessible digital agriculture applications for early disease detection.

References

Asmara, I. G. N. B. P., Kesiman, M. W. A., & Indrawan, G. (2023). Balinese Shadow Puppet Characters Detection In The Wayang Peteng Performance Using The Yolov5 Algorithm. Jurnal Nasional Pendidikan Teknik Informatika: Janapati, 12(3), 388–397.

Bitar, A., Rosales, R., & Paulitsch, M. (2023). Gradient-Based Feature-Attribution Explainability Methods For Spiking Neural Networks. Frontiers In Neuroscience, 17. Https://Doi.Org/10.3389/Fnins.2023.1153999

Dewi, C., Bilaut, F. Y., Christanto, H. J., & Dai, G. (2024). Deep Learning For The Classification Of Rice Leaf Diseases Using Yolov8.

Hameed, S., Al-Shammari, E. T., & Ramzan, N. (2026). Evaluation of deep learning models for plant disease classification using MCC and balanced accuracy metrics. Results in Engineering, 26, 111212. https://doi.org/10.1016/j.rineng.2026.111212

I Made Angga Darma Putra, Maysanjaya, I. Md. D., & Kesiman, M. W. A. (2023). Pendekatan Berbasis U-Net Untuk Segmentasi Hard Exudate Dalam Citra Fundus Retina. Insert : Information System And Emerging Technology Journal, 4(1), 26–36. Https://Doi.Org/10.23887/Insert.V4i1.59034

Lecun, Y., Bengio, Y., & Hinton, G. (2015). Deep Learning. Nature, 521(7553), 436–444.

Lecun, Y., Bottou, L., Bengio, Y., & Haffner, P. (1998). Gradient-Based Learning Applied To Document Recognition. Proceedings Of The Ieee, 86(11), 2278–2324.

Lu, J., Liu, X., Ma, X., Tong, J., & Peng, J. (2023). Improved Mobilenetv2 Crop Disease Identification Model For Intelligent Agriculture. Peerj Computer Science, 9, E1595.

Manurung, D. G., Pinasthika, M. R., Vasya, M. A. O., Putri, R. A. D. S., Tampubolon, A. P., Prayata, R. F., Nisa, S. K., & Yudistira, N. (2024). Deteksi Dan Klasifikasi Hama Potato Beetle Pada Tanaman Kentang Menggunakan Yolov8. Jurnal Teknologi Informasi Dan Ilmu Komputer, 11(4), 723–734.

Natih, I. D. G. A. W., Kesiman, M. W. A., & Sunarya, I. M. G. (2026). Analisis Perbandingan Arsitektur Dan Optimizer Yolov11 Untuk Estimasi Buah Kelapa. Riggs: Journal Of Artificial Intelligence And Digital Business, 4(4), 12–19.

Njoroge, T., Kibuku, R., & Mugoye, K. (2025). Comparative And Edge-Hybrid Modeling Of Efficientnetv2 And Mobilenetv2 For Multi-Classcrop Disease Classification With Statistical Validation. Journal Of Edge Computing, 4(2), 234–262.

Nugraha, G. S., Wijaya, I. G. P. S., Bimantoro, F., Husodo, A. Y., & Hamami, F. (2023). Arabic Character Recognition Using CNN Lenet-5. Joiv: International Journal On Informatics Visualization, 7(4), 2183–2188.

Nurokhman, A., Surorejo, S., Kurniawan, R. D., & Gunawan, G. (2024). Application Of Computer Vision Techniques To Detect Diseases And Pests Of Chili Plants. Journal Of Intelligent Decision Support System (IDSS), 7(1), 10–18.

Obafemi-Ajayi, T., Eichler, R., Walsh, B., Knisley, D., & Knisley, J. (2025). Multi-class plant disease classification using Matthews Correlation Coefficient as primary evaluation metric for imbalanced agricultural datasets. Engineering Applications of Artificial Intelligence, 139, 113347. https://doi.org/10.1016/j.engappai.2025.113347

Prasetia, I. P. W., & Sunarya, I. M. G. (2024). Image Classification Of Balinese Seasoning Base Genep Based On Deep Learning. Jurnal Nasional Pendidikan Teknik Informatika: Janapati, 13(1), 79–90.

Pusparani, D. A., Kesiman, M. W. A., & Aryanto, K. Y. E. (2024). Identification Of Little Tuna Species Using Convolutional Neural Networks (Cnn) Method And Resnet-50 Architecture. Indonesian Journal Of Artificial Intelligence And Data Mining, 8(1), 86–93.

Sandler, M., Howard, A., Zhu, M., Zhmoginov, A., & Chen, L.-C. (2018). Mobilenetv2: Inverted Residuals And Linear Bottlenecks. Proceedings Of The Ieee Conference On Computer Vision And Pattern Recognition, 4510–4520.

Saputra, I. P. G. D., Kesiman, M. W. A., & Sunarya, I. M. G. (2026). Deteksi Pemalsuan Qris Mpm Statis Menggunakan Yolo, Paddleocr Dan Metode Berbasis Aturan. Bulletin Of Computer Science Research, 6(2), 763–775.

Solichin, A., Pebrianti, D., & Riyanto, S. (2025). Modified Mobilenet-V2 Convolution Neural Network (CNN) For Character Identification Of Surakarta Shadow Puppets. In Ijacsa) International Journal Of Advanced Computer Science And Applications (Vol. 16, Number 5). Www.Ijacsa.Thesai.Org

Somvanshi, S., Islam, M. M., Chhetri, G., Chakraborty, R., Mimi, M. S., Shuvo, S. A., Islam, K. S., Javed, S., Rafat, S. A., Dutta, A., & Das, S. (2026). From Tiny Machine Learning To Tiny Deep Learning: A Survey. Acm Computing Surveys, 58(7), 1–33. Https://Doi.Org/10.1145/3776588

Sumantara, I. G. L. T., Kesiman, M. W. A., & Sunarya, I. M. G. (2024). Comparative Analysis Of Cnn Methods For Periapical Radiograph Classification. Jurnal Nasional Pendidikan Teknik Informatika: Janapati, 13(2), 204–214.

Suputra, I. P. A., Gunadi, I. G. A., & Sunarya, I. M. G. (2025). Hyperparameter Optimization With Mobilenet Architecture And Vgg Architecture For Urban Traffic Density Classification Using Bali Camera Image Data. Sinkron: Jurnal Dan Penelitian Teknik Informatika, 9(3), 1132–1145.

Tamayasa, K. A., & Dewi, L. J. E. (2026). Analisis Perbandingan Model Arsitektur Mobilenetv2 Dan Efficientnetb3 Dalam Klasifikasi Penyakit Daun Jagung. Jurnal Informatika Dan Teknik Elektro Terapan, 14(1). Https://Doi.Org/10.23960/Jitet.V14i1.8624

Terven, J., Córdova-Esparza, D.-M., & Romero-González, J.-A. (2023). A Comprehensive Review Of Yolo Architectures In Computer Vision: From Yolov1 To Yolov8 And Yolo-Nas. Machine Learning And Knowledge Extraction, 5(4), 1680–1716.

Widaningrum, I., Astuti, I. P., Mustikasari, D., Nurfitri, K., Az-Zahra, R. R., Vidyastari, R. I., Selamat, A., & Ponorogo, U. M. (2025). The Lima Pandawa Shadow Puppet Characters Utilizing Principal Component Analysis (PCA) For Feature Extraction And K-Nearest Neighbor (Knn) For Classification) Recognition Of The Lima Pandawa Shadow Puppet Characters Utilizing Principal Component Analysis (PCA) For Feature Extraction And K-Nearest Neighbor (Knn) For Classification. In Indonesian Journal Of Information Systems (IJIS) (Vol. 8, Number 1).

Wita, D. S., & Subekti, A. (2023). Mobilenet-Based Transfer Learning For Detection Of Eucalyptus Pellita Diseases. Jurnal Nasional Pendidikan Teknik Informatika: Janapati, 12(1), 1–7.

Yudhantara, I. K. P., Kertiasih, N. K., & Wahyu Wijaya, I. N. S. (2026). Perbandingan Model Klasifikasi Penyakit Daun Bakau Menggunakan Arsitektur Vgg16 Dan Mobilnetv2. Jurnal Informatika Dan Teknik Elektro Terapan, 14(1). Https://Doi.Org/10.23960/Jitet.V14i1.8816

Yu, H., Hu, X., Guo, J., Bi, C., Xue, M., & Chen, H. (2026). An Enhanced Multi-Scale Lw-Mobilenetv2 For Traditional Chinese Medicine Decoction Pieces Recognition. Biomedical Signal Processing And Control, 113, 109200. Https://Doi.Org/10.1016/J.Bspc.2025.109200