Article Open Access

Cotton Disease Prediction Using Deep Transfer Learning: Comparative Analysis of Resnet50, VGG16 and Inceptionv3 Models

Sandeep Gupta, Abu Bakar Abdul Hamid, Tadiwa Elisha Nyamasvisva, Vishal Jain, Nitin Tyagi, NG Khai Mun, Danish Ather

Abstract


Cotton is among the most critical crops in the world textile industry, but it is highly susceptible to a vast array of infections that have a tremendous impact on output and fiber quality. Traditional cotton disease diagnosis is mostly based on manual inspection by farmers and experts and is time consuming, labor intensive and inaccurate due to similarity of symptoms. The high rate at which artificial intelligence, especially computer vision and deep learning (DL), have advanced has provided effective alternatives to auto-detecting plant diseases. As a subdivision of the DL approach, transfer learning allows adapting existing convolutional neural networks to the agricultural domain using smaller datasets to guarantee higher performance. This work introduces comparative analysis of three popular deep transfer learning (DTL) models ResNet50, VGG16, and InceptionV3 that are used in the classification of cotton leaf diseases. The training, validation, and testing were performed on a dataset of 1,991 labelled images that included four categories of normal and diseased cotton leaves and plants. All models were optimized and assessed with standard measures, such as validation and test accuracy. The experimental results show that InceptionV3 had the highest accuracy of 95.28, VGG16 had 85.85, and ResNet50 had the lowest accuracy of 69.81. The high accuracy of InceptionV3 is also a testament to its ability in the extraction of multi-scale features, and the trade-off between accuracy and computational efficiency. The results affirm the feasibility of DTL frameworks to revolutionize precision agriculture by facilitating diagnosis of cotton diseases in a timely and reliable manner. This development can help in ensuring that farming activities are sustainable, pesticides are used efficiently and the economy does not suffer economic losses and helps in ensuring that productivity and environmental protection are maintained in cotton farming.


Keywords


Cotton Disease Detection, Deep Learning, Deep Transfer Learning, Resnet50, VGG16, Inception_V3

References


Islam, M. M., Talukder, M. A., Sarker, M. R. A., Uddin, M. A., Akhter, A., Sharmin, S., ... & Debnath, S. K. (2023). A deep learning model for cotton disease prediction using fine-tuning with smart web application in agriculture. Intelligent Sys-tems with Applications, 20, 200278. https://www.sciencedirect.com/science/article/pii/S2667305323001035

Sain, S. K., Paul, D., Kumar, P., Kumar, A., Mohan, M., Monga, D., ... & Prasad, Y. G. (2024). Cotton leaf curl disease (CLCuD) prediction modeling in upland cotton under different ecological conditions using machine learning tools. Ecologi-cal Informatics, 81, 102648. https://www.sciencedirect.com/science/article/pii/S1574954124001900

Zhang, N., Zhang, X., Shang, P., Ma, R., Yuan, X., Li, L., & Bai, T. (2023). Detection of cotton verticillium wilt disease severi-ty based on hyperspectrum and GWO-SVM. Remote Sensing, 15(13), 3373. https://www.mdpi.com/2072-4292/15/13/3373/pdf

Costa, E., Ajith, V., Al Khaldi, A. F., Isgrò, A., Lee, K. J., Luigetti, R., ... & Thirstrup, S. (2025). Addressing global regulatory challenges in rare disease drug development. Drug Discovery Today, 104462. https://www.sciencedirect.com/science/article/pii/S1359644625001758

Kang, X., Huang, C., Zhang, L., Yang, M., Zhang, Z., & Lyu, X. (2023). Assessing the severity of cotton Verticillium wilt dis-ease from in situ canopy images and spectra using convolutional neural networks. The Crop Journal, 11(3), 933-940. https://www.sciencedirect.com/science/article/pii/S221451412200263X

Bishshash, P., Nirob, A. S., Shikder, H., Sarower, A. H., Bhuiyan, T., & Noori, S. R. H. (2024). A comprehensive cotton leaf disease dataset for enhanced detection and classification. Data in Brief, 57, 110913. https://www.sciencedirect.com/science/article/pii/S235234092400876X

Xu, B., Yang, J., Chen, D., Wang, X., Ai, X., Liu, L., ... & Fan, L. (2025). Screening Verticillium wilt-resistant germplasm by monitoring the time-series chlorophyll content of cotton canopies via a UAV-based high-throughput platform. Computers and Electronics in Agriculture, 238, 110791. https://www.sciencedirect.com/science/article/pii/S016816992500897X

Abdalla, A., Wheeler, T. A., Dever, J., Lin, Z., Arce, J., & Guo, W. (2024). Assessing fusarium oxysporum disease severity in cotton using unmanned aerial system images and a hybrid domain adaptation deep learning time series model. Biosys-tems Engineering, 237, 220-231. https://par.nsf.gov/servlets/purl/10615176

Quddus, M. A., Chowdhury, S., Jasper, W., & Das, L. (2025). Enhancing cotton yield prediction with robust deep neural network-based framework. European Journal of Agronomy, 170, 127750. https://www.sciencedirect.com/science/article/pii/S1161030125002461

Cao, B., Zhou, P., Chen, W., Wang, H., & Liu, S. (2024). Real-Time Monitoring and Early Warning of Cotton Diseases and Pests Based on Agricultural Internet of Things. Procedia Computer Science, 243, 253-260. https://www.sciencedirect.com/science/article/pii/S1877050924020398/pdf?md5=2a551f2811a5b3dedc73593b3e03e12d&pid=1-s2.0-S1877050924020398-main.pdf

Tan, F., Cang, H., Gao, X., Wu, N., Di, R., Zhang, Y., ... & Zhang, C. (2025). Early detection of cotton Verticillium wilt based on generative adversarial networks and hyperspectral imaging technology. Industrial Crops and Products, 231, 121167. https://www.sciencedirect.com/science/article/pii/S0926669025007137

Shahzad, M. U., Tahir, S., Rashid, J., Khashan, O. A., Ahmad, R., Mansoor, S., & Ghani, A. (2025). Machine Learning-Based Cotton Yield Forecasting under Climate Change for Precision Agriculture. Smart Agricultural Technology, 101117. https://www.sciencedirect.com/science/article/pii/S2772375525003508

Wang, T., Mei, X., Thomasson, J. A., Yang, C., Han, X., Yadav, P. K., & Shi, Y. (2022). GIS-based volunteer cotton habitat prediction and plant-level detection with UAV remote sensing. Computers and Electronics in Agriculture, 193, 106629. https://www.sciencedirect.com/science/article/am/pii/S0168169921006463

Cotton, A., Salerno, P. R., Chen, Z., Virani, S., Sattar, N., Rajagopalan, S., & Deo, S. V. (2025). Forecasting Atherosclerotic Cardiovascular Disease in South Asia Until 2040: A Bayesian Modeling Approach. JACC: Asia. https://www.sciencedirect.com/science/article/pii/S2772374725003667

Tian, Y., Jiang, Y., Zeng, M., Hui, J., & Jiang, Q. (2025). Inversion of cotton leaf area index under Verticillium wilt stress from UAV multispectral images: Deep learning-based vs. classical-based algorithms. Industrial Crops and Products, 233, 121345. https://www.sciencedirect.com/science/article/pii/S092666902500891X

Madu, A., Alex-Okoro, T., Okoduwa, A., & Cotton, J. (2024). Austrian syndrome: Report of one case and a systematic re-view of case reports–new insights. Clinical Medicine, 24(3), 100205. https://www.sciencedirect.com/science/article/pii/S1470211824053909

Yu, G., Ma, B., Zhang, R., Xu, Y., Lian, Y., & Dong, F. (2025). CPD-YOLO: A cross-platform detection method for cotton pests and diseases using UAV and smartphone imaging. Industrial Crops and Products, 234, 121515. https://www.sciencedirect.com/science/article/pii/S0926669025010611

Pudumalar, S., & Muthuramalingam, S. (2024). Hydra: An ensemble deep learning recognition model for plant diseases. Journal of Engineering Research, 12(4), 781-792. https://www.sciencedirect.com/science/article/pii/S2307187723002584

Hussain, S., Ghramh, H. A., Rafiq, M. S., Sneharani, A. H., Shah, S. M. A., Ullah, M. I., ... & Mahmood, K. (2023). Tempera-ture-based prediction and validation of pink bollworm, Pectinophora gossypiella infestation on cotton crop. Journal of King Saud University-Science, 35(2), 102494. https://www.sciencedirect.com/science/article/pii/S1018364722006759

Huo, W. Q., Zhang, Z. Q., Ren, Z. Y., Zhao, J. J., Song, C. X., Wang, X. X., ... & Ma, X. F. (2023). Unraveling genomic regions and candidate genes for multiple disease resistance in upland cotton using meta-QTL analysis. Heliyon, 9(8). https://www.cell.com/heliyon/fulltext/S2405-8440(23)05939-X?uuid=uuid%3Ac1274ef8-7cf6-4c00-a382-05fd934cde8c




DOI: https://doi.org/10.52088/ijesty.v5i3.1116

Refbacks

  • There are currently no refbacks.


Copyright (c) 2025 Sandeep Gupta, Abu Bakar Abdul Hamid, Tadiwa Elisha Nyamasvisva, Vishal Jain, Nitin Tyagi, NG Khai Mun, Danish Ather

International Journal of Engineering, Science, and Information Technology (IJESTY) eISSN 2775-2674