Rice leaf diseases prediction using deep neural networks with transfer learning

Area of Science:

• Agricultural Science
• Computer Vision
• Deep Learning

Background:

• Rice is a global staple food, crucial for energy. Rice yield is threatened by various biotic and abiotic factors, including diseases.
• Current disease detection methods rely on manual inspection, which is time-consuming, resource-intensive, and prone to inaccuracies.

Purpose of the Study:

• To develop an automated system for accurate and efficient identification of rice plant diseases.
• To leverage deep learning, specifically convolutional neural networks (CNNs), for image-based disease diagnosis in rice.

Main Methods:

• Utilized the InceptionResNetV2 architecture, a type of CNN, for image classification.
• Employed a transfer learning approach to optimize the model for recognizing diseases in rice leaf images.
• Trained and evaluated the model on a dataset of rice leaf images to assess its classification performance.

Main Results:

• The InceptionResNetV2 model achieved a high accuracy of 95.67% in classifying rice leaf diseases.
• The optimized model demonstrated effective performance in distinguishing between healthy and diseased rice plant images.

Conclusions:

• Deep learning models, such as InceptionResNetV2, show significant potential for automated disease detection in agriculture.
• This approach offers a more efficient and accurate alternative to traditional manual methods, supporting sustainable rice farming practices.