Pathogen Detection in Dragonfruit With Transfer Learning and Fine-Tuned Keras Models

Abstract

Dragonfruit, also known as pitaya, is a tropical fruit vulnerable to various diseases that can significantly impact its growth and yield. Timely and accurate disease detection and classification are crucial for effective management of dragonfruit cultivation. Transfer learning, a powerful machine learning technique, has shown promise in plant disease classification by leveraging pre-trained models. Keras, a popular deep learning library, offers pre-trained models suitable for image classification tasks. This research aims to develop a robust classification model for accurately identifying and distinguishing diseases by considering self-created dragonfruit plant dataset. By fine-tuning pre-trained Keras models, specifically adapted for dragonfruit disease classification, transfer learning is employed to achieve high accuracy in disease identification. The study contributes to agricultural disease detection by empowering farmers and researchers with a valuable tool for disease monitoring and management.