Deep Learning-Based Approach for Plant Disease Detection and Classification
International Journal of Research and Review in Applied Science, Humanities, and Technology · 28.12.2024 · 234-242
DOI: 10.71143/ftfma695 ↗Kurz gefasst
Fachartikel in der Zeitschrift International Journal of Research and Review in Applied Science, Humanities, and Technology, online erschienen am 28.12.2024. Verfasst von Pooja Sharma und 3 weiteren Autor:innen. Der Volltext ist frei zugänglich (Open Access, CC BY-NC 4.0).
Abstract (Originalsprache)
Plant diseases are a major cause of yield loss and food insecurity, particularly in agrarian economies where expert diagnosis is not available at the farm level. Manual inspection is slow, subjective and difficult to scale. This paper presents a deep learning-based approach for automatic detection and classification of plant diseases from leaf images. The proposed model uses an ImageNet-pretrained EfficientNet-B0 backbone combined with a squeeze-and-excitation (SE) channel-attention block and a lightweight classification head, trained using a two-stage transfer-learning strategy with extensive data augmentation. The model is evaluated on the PlantVillage dataset containing 54,305 images across 38 classes of healthy and diseased leaves from 14 crop species. The proposed model achieves a test accuracy of 99.18%, macro precision of 99.05%, macro recall of 98.97% and macro F1-score of 99.01% with only 4.4 million parameters and an average inference time of 7.8 ms per image. It outperforms a baseline CNN, VGG16, ResNet50 and MobileNetV2 trained under identical conditions. An ablation study confirms the contribution of each component, and a cross-dataset test on field images from PlantDoc highlights the generalization gap and the benefit of limited field fine-tuning. The results show that the approach is accurate and light enough for deployment on mobile and edge devices for real-time crop health monitoring.
Zitieren
APA
Sharma, P., Anisha, Jha, R., & Gupta, S. (2024). Deep Learning-Based Approach for Plant Disease Detection and Classification. International Journal of Research and Review in Applied Science, Humanities, and Technology, 234-242. https://doi.org/10.71143/ftfma695
Vancouver
Sharma P, Anisha, Jha R, Gupta S. Deep Learning-Based Approach for Plant Disease Detection and Classification. International Journal of Research and Review in Applied Science, Humanities, and Technology. 2024:234-242. doi:10.71143/ftfma695
BibTeX
@article{Sharma2024,
author = {Sharma, Pooja and Anisha and Jha, Roshni and Gupta, Sorabh},
title = {Deep Learning-Based Approach for Plant Disease Detection and Classification},
journal = {International Journal of Research and Review in Applied Science, Humanities, and Technology},
year = {2024},
pages = {234--242},
doi = {10.71143/ftfma695}
}
EUVIMED