ARCHIVES
VOL. 11, ISSUE 12 (2024)
Deploying mobilenet for efficient and real-time identification of castor leaf diseases using smartphone applications
Authors
Epaphras Kyomnom Peter, Saratu Habu
Abstract
The increasing need for sustainable
agriculture has necessitated the development of automated systems for plant
disease detection. This study presents a robust deep learning approach to
classify Castor (Ricinus communis) leaf diseases using high-resolution
images. A dataset comprising 930 images, representing six categories (Healthy,
Seedling Blight, Leaf Blight, Rust, Brown Leaf Spot, and Bacterial Leaf Spot),
was collected across diverse environmental conditions in Nigeria using a Google
Pixel 6 Pro smartphone. Data preprocessing techniques, including resizing,
normalization, and augmentation, were applied to enhance model generalization.
A MobileNet architecture, fine-tuned with pre-trained ImageNet weights, was
utilized due to its computational efficiency and suitability for mobile deployment.
The model achieved an exceptional overall accuracy of 96.6%, with precision,
recall, and F1 scores exceeding 86.7% across all disease categories. Minor
misclassifications occurred between visually similar diseases, such as Rust and
Bacterial Leaf Spot. The trained model was optimized using TensorFlow Lite and
integrated into a Flutter-based mobile application, enabling real-time disease
detection on smartphones. The app processes leaf images locally, ensuring
privacy, and provides actionable recommendations to aid timely interventions.
The system’s high accuracy and mobile compatibility demonstrate its potential
for large-scale adoption, empowering farmers with an accessible and efficient
tool for Castor crop disease management. This research highlights the efficacy
of lightweight deep learning models and mobile technology in advancing
precision agriculture, particularly in resource-limited settings.
Download
Pages:74-79
How to cite this article:
Epaphras Kyomnom Peter, Saratu Habu "Deploying mobilenet for efficient and real-time identification of castor leaf diseases using smartphone applications". International Journal of Multidisciplinary Research and Development, Vol 11, Issue 12, 2024, Pages 74-79
Download Author Certificate
Please enter the email address corresponding to this article submission to download your certificate.