Comprehensive Review of Deep Learning Methods for Identifying Okra Plant Diseases
Volume 20, Issue 1, 2026
Download| Author(s): |
Saba Yousha* The Begum Nusrat Bhutto Women University, Sukkur , Pakistan, saba.yousha@bnbwu.edu.pk Syed Atir Iftikhar IQRA University, Gulshan e Iqbal Campus, Karachi, Pakistan, syed.atir@iqra.edu.pk Ram Chand The Begum Nusrat Bhutto Women University, Sukkur , Pakistan, ram.chand@bnbwu.edu.pk Hira Kalwar The Begum Nusrat Bhutto Women University, Sukkur , Pakistan, hira.kalwar@bnbwu.edu.pk Shahzad Nasim The Begum Nusrat Bhutto Women University, Sukkur , Pakistan, shahzad.nasim@bnbwu.edu.pk |
|---|---|
| Abstract | Plant diseases and bug pests must be detected quickly, accurately, and efficiently due to the growing requirement for responsible agricultural production. Traditional methods of disease tracking and pest detection are labor-intensive, time-consuming, expensive, and challenging to scale across broad farming regions since they primarily rely on manual field inspection and specialist knowledge. Systematic sickness and identifying insects through visual analysis has been made possible by recent developments in artificial intelligence, and especially the use of deep learning. A thorough analysis of deep learning-based methods for the diagnosis, characterization, and classification of pests in farming and crop diseases is presented in this work. Including popular architectures like Convolutional Neural Networks (CNNs), YOLO-based models, Faster R-CNN, SSD, U-Net, Vision Transformers, and other cutting-edge architectures, the study methodically investigates key AI tasks like object recognition, image classification, lexical and instance segmentation, and change detection. The review also examines how large-scale already trained models, data augmenting, and transfer learning in general might enhance model effectiveness and lower training requirements. Several publicly available and field-gathered information sets, scanning settings, and commonly used criteria for evaluation are also discussed in order to provide comparisons of current approaches. Particular attention is paid to the challenges of everyday agriculture, including variations in lighting and conditions, complex conditions, obstruction, small and advanced lesions, small and unbalanced datasets, cross-class similarity, and limited modeling generalization based across various crops and spatial geographic areas. The paper also identifies significant research gaps related to dataset diversity, model accessibility, computational effectiveness, real-time setup, and the capacity for transfer of algorithms from lab into field environments. |
| Keywords | deep learning, disease detection, convolutional neural network (CNN), pest identification, image processing |
| Year | 2026 |
| Volume | 20 |
| Issue | 1 |
| Type | Research paper, manuscript, article |
| Journal Name | Journal of Information & Communication Technology | Publisher Name | ILMA University | Jel Classification | - | DOI | - | ISSN no (E, Electronic) | 2075-7239 | ISSN no (P, Print) | 2415-0169 | Country | Pakistan | City | Karachi | Institution Type | University | Journal Type | Open Access | Manuscript Processing | Blind Peer Reviewed | Format | Paper Link | https://jict.ilmauniversity.edu.pk/journal/jict/20.1/2.pdf | Page | 11-21 |