Nagineni Venkata Sireesha; Gillala Rekha

Comparative Evaluation of Fine-Tuned Transfer Learning CNN Architectures for Automated Citrus Fruit Disease Classification

PDF (2701KB), PP.1-33

Views: 0 Downloads: 0

Author(s)

Nagineni Venkata Sireesha ¹ Gillala Rekha ^1,*

1. Department of Computer Science and Engineering, Koneru Lakshmaiah Education Foundation, KLEF Deemed to be University, Aziz Nagar, 500075, Hyderabad, Telangana, India

* Corresponding author.

DOI: https://doi.org/10.5815/ijem.2026.05.01

Received: 21 May 2026 / Revised: 15 Jun. 2026 / Accepted: 10 Sep. 2026 / Published: 8 Oct. 2026

Index Terms

Citrus disease detection, Transfer learning, Deep learning, CNN, Image classification, Precision agriculture

Abstract

Early and accurate detection of citrus diseases is essential for maintaining fruit quality, minimising crop losses, and supporting sustainable agricultural production. Automated image-based diagnostic systems offer a scalable alternative to conventional manual inspection, which is often time-intensive, subjective, and susceptible to diagnostic variability. This study presents a transfer-learning-based deep learning framework for automated classification of citrus fruit diseases using a curated image collection derived from Food and Agriculture Organisation (FAO) resources. The dataset was organised into two classification groups: lemon and orange. The lemon dataset initially comprised 208 images across four classes: Healthy, Canker, Mold, and Scab, while the orange dataset contained 2,240 images across two classes. To address data scarcity and improve model generalisation, targeted data augmentation involving zooming (0.8–1.2), rotation (35°), and horizontal flipping was employed, increasing the datasets to 1,600 and 4,000 images for lemon and orange, respectively. Five ImageNet-pretrained convolutional neural network architectures—VGG16, ResNet50, InceptionV3, DenseNet121, and EfficientNetB0—were fine-tuned and evaluated using a stratified 70:30 training–testing protocol with five-fold cross-validation. Model performance was assessed using accuracy, precision, recall, and F1-score under a standardized experimental configuration. The results demonstrate that InceptionV3 achieved the highest classification accuracy of 90.0% on the four-class lemon dataset, while DenseNet121 obtained the best accuracy of 93.0% on the binary orange dataset. These findings indicate that appropriate transfer learning and targeted augmentation can substantially improve classification performance and model generalisation, particularly in limited-data agricultural imaging scenarios. The proposed benchmarking framework enables systematic, controlled comparisons of fine-tuned CNN architectures under consistent preprocessing, augmentation, training, and evaluation conditions. The resulting models can be integrated into real-time citrus disease diagnostic systems and deployed on resource-constrained platforms, including mobile and edge-computing devices, supporting scalable precision-agriculture applications.

Cite This Paper

Nagineni Venkata Sireesha, Gillala Rekha, "Comparative Evaluation of Fine-Tuned Transfer Learning CNN Architectures for Automated Citrus Fruit Disease Classification", International Journal of Engineering and Manufacturing (IJEM), Vol.16, No.5, pp.1-33, 2026. DOI:10.5815/ijem.2026.05.01

Reference

[1]Li, L., S. Zhang, and B. Wang. "Plant Disease Detection and Classification by Deep Learning-A Review." IEEE Access, vol. 9, 2021, pp. 56683–56698. https://doi.org/10.1109/ACCESS.2021.3069646.
[2]Shoaib, Muhammad, et al. "Leveraging Deep Learning for Plant Disease and Pest Detection: A Comprehensive Review and Future Directions." Frontiers in Plant Science, vol. 16, 2025. https://doi.org/10.3389/fpls.2025.1538163.
[3]Zheng, Y.-Y., et al. "CropDeep: The Crop Vision Dataset for Deep-Learning-Based Classification and Detection in Precision Agriculture." Sensors, vol. 19, no. 5, 2019, p. 1058. https://doi.org/10.3390/s19051058.
[4]Dananjayan, Sathian, et al. "Assessment of state-of-the-art deep learning-based citrus disease detection techniques using annotated optical leaf images." Computers and Electronics in Agriculture, vol. 193, 2022, 106658. https://doi.org/10.1016/j.compag.2021.106658.
[5]Devora-Guadarrama, M., et al. "Deep Learning-Based Citrus Canker and Huanglongbing Disease Detection Using Leaf Images." Computers, vol. 14, no. 11, 2025, p. 500. https://doi.org/10.3390/computers14110500.
[6]Zhang, Xinxing, et al. "Automated identification of citrus diseases in orchards using deep learning." Biosystems Engineering, vol. 223, 2022, pp. 249–258. https://doi.org/10.1016/j.biosystemseng.2022.09.006.
[7]R. M., S., et al. "A Multi-kernel CNN model with attention mechanism for classification of citrus plant diseases." Sci Rep, vol. 15, 2025, p. 24047. https://doi.org/10.1038/s41598-025-08557-3.
[8]Zhu, H., et al. "YOLOV8-CMS: a high-accuracy deep learning model for automated citrus leaf disease classification and grading." Plant Methods, vol. 21, 2025, p. 88. https://doi.org/10.1186/s13007-025-01396-3.
[9]Barman, Utpal, and Ridip Dev Choudhury. "Smartphone-assisted deep neural network to detect citrus diseases in Agri-informatics." Global Transitions Proceedings, vol. 3, no. 2, 2022, pp. 392–398. https://doi.org/10.1016/j.gltp.2021.10.004.
[10]Pal, C., et al. "A lightweight and explainable CNN model for empowering plant disease diagnosis." Sci Rep, vol. 15, 2025, p. 30720. https://doi.org/10.1038/s41598-025-94083-1.
[11]Hafza Eman, et al. "CitriBEiTNet: A Hybrid CNN-Transformer Architecture Combining MobileNetV2 with BEiT’s Global Attention for Automated Citrus Leaf Disease Diagnosis." bioRxiv, 2025. https://doi.org/10.64898/2025.12.09.693306.
[12]Karthik, R., et al. "A dual track deep fusion network for citrus disease classification using group shuffle depthwise feature pyramid and Swin transformer." Ecological Informatics, vol. 78, 2023, 102302. https://doi.org/10.1016/j.ecoinf.2023.102302.
[13]Sireesha, N. V., et al. "Unassailable citrus disease classification via multi-stage deep ensemble learning with vision transformers." Sci Rep, vol. 15, 2025, p. 40573. https://doi.org/10.1038/s41598-025-24416-7.
[14]Sireesha, N. V., et al. "Interpretable citrus disease classification through GradCAM++ LIME and vision transformer-based deep learning." Discov Appl Sci, 2026. https://doi.org/10.1007/s42452-026-08292-y.
[15]Yadav, Pappu Kumar, et al. "Automated Classification of Citrus Disease on Fruits and Leaves Using Convolutional Neural Network Generated Features from Hyperspectral Images and Machine Learning Classifiers." Journal of Applied Remote Sensing, vol. 18, no. 1, 2024, p. 014512. https://doi.org/10.1117/1.JRS.18.014512.
[16]Min, Hyun Jung, et al. "Classification of Citrus Leaf Diseases Using Hyperspectral Reflectance and Fluorescence Imaging and Machine Learning Techniques." Horticulturae, vol. 10, no. 11, 2024, p. 1124. https://doi.org/10.3390/horticulturae10111124.
[17]Li, Yuchen, et al. "Denoising Diffusion Probabilistic Models and Transfer Learning for Citrus Disease Diagnosis." Frontiers in Plant Science, vol. 14, 2023. https://doi.org/10.3389/fpls.2023.1267810.
[18]Lanjewar, M. G., and J. S. Parab. "CNN and transfer learning methods with augmentation for citrus leaf diseases detection using PaaS cloud on mobile." Multimed Tools Appl, vol. 83, 2024, pp. 31733–31758. https://doi.org/10.1007/s11042-023-16886-6.
[19]Kujur, L., Gupta, V., and A. Singhal. "A hybrid multi-optimiser approach using CNN and GB for accurate prediction of citrus fruit diseases." Discover Applied Sci, vol. 7, 2025, p. 151. https://doi.org/10.1007/s42452-025-06593-2.
[20]Zhong, Fangming, et al. "Zero- and few-shot learning for diseases recognition of Citrus aurantium L. using conditional adversarial autoencoders." Computers and Electronics in Agriculture, vol. 179, 2020, 105828. https://doi.org/10.1016/j.compag.2020.105828.
[21]Ye, N., et al. "Early detection of Citrus Huanglongbing by UAV remote sensing based on MGA-UNet." Front. Plant Sci., vol. 16, 2025. https://doi.org/10.3389/fpls.2025.1503645.
[22]P. Dhiman, A. Kaur, Y. Hamid, E. Alabdulkreem, H. Elmannai, and N. Ababneh, “Smart disease detection system for citrus fruits using deep learning with edge computing,” Sustainability, vol. 15, no. 5, Art. no. 4576, Mar. 2023, https://doi.org/10.3390/su15054576.
[23]S. Faisal, K. Javed, S. Ali, A. Alasiry, M. Marzougui, M. A. Khan, and J.-H. Cha, "Deep transfer learning based detection and classification of citrus plant diseases," Computers, Materials & Continua, vol. 76, no. 1, pp. 895–914, 2023. doi: 10.32604/cmc.2023.039781. https://doi.org/10.32604/cmc.2023.039781.
[24]A. Goyal and K. Lakhwani, "Integrating advanced deep learning techniques for enhanced detection and classification of citrus leaf and fruit diseases," Scientific Reports, vol. 15, no. 1, Art. no. 12659, 2025. doi: 10.1038/s41598-025-97159-0. https://doi.org/10.1038/s41598-025-97159-0.
[25]N. Butt, M. M. Iqbal, S. Ramzan, A. Raza, L. Abualigah, N. L. Fitriyani, Y. Gu, and M. Syafrudin, "Citrus diseases detection using innovative deep learning approach and Hybrid Meta-Heuristic," PLoS ONE, vol. 20, no. 1, Art. no. e0316081, Jan. 2025. https://doi.org/10.1371/journal.pone.0316081.

International Journal of Engineering and Manufacturing (IJEM)