DeepCrop: Deep learning-based crop disease prediction with web application

• Published in: Journal of agriculture and food research 2023-12, Vol.14, p.100764, Article 100764
• Main Authors: Islam, Md. Manowarul, Adil, Md Abdul Ahad, Talukder, Md. Alamin, Ahamed, Md. Khabir Uddin, Uddin, Md Ashraf, Hasan, Md. Kamran, Sharmin, Selina, Rahman, Md. Mahbubur, Debnath, Sumon Kumar
• Format: Article
• Language: English
• Subjects: Agriculture; Crop; Disease; Farmers; Plant; Transfer Learning; Treatment
• ISSN: 2666-1543; 2666-1543
• DOI: 10.1016/j.jafr.2023.100764

Agriculture plays a significant role in every nation's economy by producing crops. Plant disease identification is one of the most important aspects of maintaining an agriculturally developed nation. The timely and efficient detection of plant diseases is essential for a healthy and productive agricultural sector and to prevent wasting money and other resources. Various diseases that could affect a plant cause crop farmers to lose a substantial sum yearly. Deep learning can play a crucial role in helping farmers prevent crop failure by early disease detection in plant leaves. In the experiment, we examined CNN, VGG-16, VGG-19 and ResNet-50 models on plant-village 10000 image dataset to detect crop infection and got the accuracy rate of 98.60%, 92.39%, 96.15%, and 98.98% for CNN, VGG-16, VGG-19 and ResNet-50 respectively. The study indicates that ResNet-50 outperforms the other models with an accuracy of 98.98%. So, the ResNet50 model was chosen to be developed into a smart web application for real-life crop disease prediction. The proposed web application aims to assist farmers in identifying diseases of plants by analyzing photos of the plant leaves. The proposed application uses the ResNet50 transfer learning model at its heart to distinguish healthy and infected leaves and classify the present disease type. The goal is to help farmers save resources and prevent economic loss by detecting plant diseases early and applying the appropriate treatment. [Display omitted] •Proposed a deep learning-based model for crop disease detection.•Provides a higher accuracy rate of 98.98% using ResNet-50 for disease detection.•Ensure farmers save resources and prevent economic loss.