Utilization of the Convolutional Neural Network Method for Detecting Banana Leaf Disease

Utilization of the Convolutional Neural Network Method for Detecting Banana Leaf Disease

agricultural sector;artificial Intelligence;Banana leaf disease; CNN Model; technology development

anana leaf diseases such as Sigatoka, Cordana, and Pestalotiopsis pose a significant threat to banana productivity, with implications for food security and the global economy. Early detection of this disease is an important step to reduce its spread and maintain crop yield stability. This research utilizes the Convolutional Neural Network (CNN) method to detect banana leaf diseases based on image analysis of infected and healthy leaves. The dataset used includes 937 images consisting of four maincategories, namely healthy leaves, Sigatoka, Cordana, and Pestalotiopsis. The dataset is processed through augmentation to increase data diversity and quality. The CNN model was applied for classification, with evaluation results reaching 92.85% accuracy, 95.73% recall, 93.52% precision, and 94.60% F1-score. This research contributes to the development of Artificial Intelligence-based technology for applications in the agricultural sector, especially in supporting farmers to detect banana leaf diseases quickly, accurately and efficiently. The research results also provide recommendations for exploring additional data augmentation and increasing dataset variety to improve model detection performance in the future. This shows CNN's potential to supportsustainable agriculture in the modern era

Nita Helmawati1*, Ema Utami

https://jurnal.iaii.or.id/index.php/RESTI/article/view/6140/996

Magister of Informatics, Universitas Amikom Yogyakarta, Yogyakarta, Indonesia

28-12-2024

FAJAR BAGUS W

PDF

ENGLISH

TEXT