Anthurium Leaf Diseases Identification Using Deep Learning

Tropical ornamental plants such as anthurium are cultivated for their attractive foliage and long-lasting flowers and have gained importance as export-quality cut flowers. Healthy growth and high flower yield depend on suitable physical and chemical conditions of the growing medium and the use of high-quality planting materials. However, anthurium cultivation is significantly affected by plant diseases, particularly due to difficulties in accurate disease identification and limited farmer awareness. Some diseases, such as bacterial infections, have no effective cure and must be detected early to prevent widespread damage.
This study presents a deep learning–based approach for identifying anthurium leaf diseases using Convolutional Neural Networks (CNNs). CNNs are powerful imageprocessing models capable of extracting complex features from leaf images while efficiently utilizing GPU-based computation. Pre-trained CNN models were employed to classify images of healthy anthurium leaves and two major bacterial diseases—bacterial blight and bacterial wilt.
Approximately 400 leaf images were used for training and testing the models. Model performance was evaluated by adjusting key hyperparameters, including the number of epochs, batch size, and learning rate. The optimal configuration consisted of 25 training epochs, a batch size of 32, and a learning rate of 0.001, achieving a test accuracy of 98.44%. Further evaluation using advanced deep learning architectures, namely ResNet-50 and VGG-16, resulted in classification accuracies of 100%, demonstrating the effectiveness of deeper network structures.
The final model was implemented in an Android mobile application that enables disease detection using images captured via a mobile camera or image gallery. The high accuracy and practical deployment of the system highlight its potential as a reliable tool for early disease diagnosis, supporting farmers in effective disease management and promoting healthier anthurium cultivation.