Loading article...
Loading article...
Stream/Area:
-
Subject:
-
Paper ID:
IJSREM69018
Volume:
10 |
11 |
Nov 2026
Pages:
1-10
Publication Frequency:
Monthly (12 issues per year)
License:International Journal of Scientific Research in Engineering and Management follows the Creative Commons Attribution 4.0 (CC BY 4.0) License, allowing free use and distribution with proper attribution.

Information Technology
PCET's Nutan Maharashtra Institute of Engineering and Technology, Pune

Information Technology
PCET's Nutan Maharashtra Institute of Engineering and Technology, Pune

Information Technology
PCET's Nutan Maharashtra Institute of Engineering and Technology, Pune

Information Technology
PCET's Nutan Maharashtra Institute of Engineering and Technology, Pune
Crop diseases reduce agricultural productivity and can threaten food quality and farm income. Image-based diagnosis using convolutional neural networks (CNNs) has become an active research direction because CNNs learn visual features from labelled leaf images. This review examines CNN-based crop leaf disease classification with emphasis on dataset characteristics, architecture selection, transfer learning, preprocessing, evaluation design and deployment constraints. It discusses conventional CNNs, VGG, residual networks, MobileNet, EfficientNet and DenseNet, notes transformer-based and hybrid alternatives, and considers the different roles of controlled datasets such as PlantVillage and field-oriented collections such as PlantDoc. Published studies report high accuracy on curated benchmarks, but results obtained under different class definitions, image sources and split protocols are not directly comparable. Evidence that a model performs well on a controlled test set is not, by itself, evidence of reliable farm-level diagnosis. The review therefore analyses dataset shift, class imbalance, duplicate-image leakage, label quality, explainability, uncertainty estimation and mobile inference. It proposes a practical evidence hierarchy progressing from internal random splits to grouped evaluation, cross-dataset testing and prospective field assessment. The central conclusion is that model selection should balance predictive performance with generalisation, computational cost and the ability to recognise uncertain or unsupported inputs.
KEYWORDS
CNN-based methods have become an important approach to crop leaf disease classification. Conventional CNNs, VGG, ResNet, MobileNet, EfficientNet and DenseNet provide different trade-offs in feature learning, computational cost and deployment suitability. Transfer learning can make these architectures practical when labelled agricultural data are limited, while preprocessing and augmentation can improve robustness when used carefully. Public datasets such as PlantVillage have enabled large-scale benchmarking, whereas field-oriented resources such as PlantDoc expose additional visual complexity. The central conclusion of this review is that high accuracy on controlled images is not sufficient evidence of reliable agricultural deployment. Dataset realism, leakage prevention, class-wise performance, external validation, explainability, calibration and mobile efficiency should be considered together. Model selection should be guided by the intended use case and the quality of evidence available, not by a single headline accuracy value. Future work should prioritise field data, transparent evaluation and human-centred decision support. CNN classifiers can assist crop-health monitoring, but their predictions should be interpreted within the limits of the training data and confirmed by appropriate agricultural expertise when necessary.
[1] S. P. Mohanty, D. P. Hughes, and M. Salathé, “Using deep learning for image-based plant disease detection,” Frontiers in Plant Science, vol. 7, Art. no. 1419, 2016. [2] K. P. Ferentinos, “Deep learning models for plant disease detection and diagnosis,” Computers and Electronics in Agriculture, vol. 145, pp. 311–318, 2018. [3] D. Singh et al., “PlantDoc: A dataset for visual plant disease detection,” in Proc. 7th ACM IKDD CoDS and 25th COMAD, 2020, pp. 249–253. [4] K. He, X. Zhang, S. Ren, and J. Sun, “Deep residual learning for image recognition,” in Proc. IEEE CVPR, 2016, pp. 770–778. [5] A. G. Howard et al., “MobileNets: Efficient convolutional neural networks for mobile vision applications,” arXiv:1704.04861, 2017. [6] M. Tan and Q. V. Le, “EfficientNet: Rethinking model scaling for convolutional neural networks,” in Proc. ICML, 2019, pp. 6105–6114. [7] D. P. Hughes and M. Salathé, “An open access repository of images on plant health to enable the development of mobile disease diagnostics,” arXiv:1511.08060, 2015. [8] I. Pacal et al., “A systematic review of deep learning techniques for plant diseases,” Artificial Intelligence Review, vol. 57, Art. no. 304, 2024. [9] A. Upadhyay et al., “Deep learning and computer vision in plant disease detection: A comprehensive review,” Artificial Intelligence Review, vol. 58, Art. no. 92, 2025. [10] E. C. Too et al., “A comparative study of fine-tuning deep learning models for plant disease identification,” Computers and Electronics in Agriculture, vol. 161, pp. 272–279, 2019.
Publication Date
11/16/2026
Peer Review Type
Double-Blind Peer Review
Review Rounds: 2
Documents
Article Views
0
Total views since publication
PDF Downloads
3
Total downloads since publication
Geographic Reach
5
Digital Preservation: This article is permanently archived on Zenodo and complies with all data preservation standards across major indexing platforms.
Plagiarism Screening
This manuscript has been checked for plagiarism using TurnitinPlagiarismChecker software.
Conflict of Interest
The authors declare no conflict of interest related to this research work.
Funding Information
No funding information available for this research.
Related research in the same subject area
Loading similar articles...