Cotton plant disease-prediction using Image processing and Transfer learning

Abstract: This paper tackles the challenge of identifying diseases in cotton plants using deep-learning techniques. The aim of using convolutional neural networks (CNNs) is to categorize five photos of cotton leaves into groups, such as healthy and unhealthy. Implemented with Keras and TensorFlow frameworks, the paper offers a detailed workflow from data collection and pre-processing to model training and evaluation. The dataset includes images of cotton leaves, with data augmentation techniques enhancing the training process to guarantee precise categorization, the CNN architecture takes significant elements out of the pictures. Another strategy being investigated to boost performance with little data is transfer learning. Transfer learning is also explored to improve performance with limited data. Evaluation metrics like accuracy, precision, recall, and F1-score are used to assess the model. This project serves as both an educational resource and a practical tool for agricultural stakeholders, promoting early disease detection, better crop management, and improved yield through the application of AI in agriculture.

Keywords: Deep learning, Convolutional Neural Networks (CNNs), Image classification, Keras, TensorFlow, Data augmentation, Transfer learning