Integration of Capsule Network with CNN for Plant Leaf Disease Classification

Abstract

Plant leaf disease classification is challenging due to the wide variation in disease symptoms and the diverse morphological characteristics of plant leaves. These variations complicate model training and hinder classification accuracy. This study proposed a hybrid deep learning (DL) model for leaf disease training and classification. The proposed model integrates Capsule Networks (CN) for spatial relationship retention, SE-Residual blocks improve feature extraction while minimizing information loss, and CN capture spatial relationships with reduced dependency on large datasets, and Long Short-Term Memory (LSTM) to enhance training efficiency. The proposed model was trained and evaluated using the Rice Leaf Disease Dataset (RLDD). Its performance was compared with existing state-of-the-art models. The experimental results showed that the proposed model achieved the highest training accuracy of 96.01%, classification results 75.67% for bacterial leaf blight, 80.43% for brown spot, 86.67% for healthy, 76.52% for leaf blight, 98.96% for leaf scald, and 93.18% for narrow brown spot. These results highlight the effectiveness of the proposed model in achieving high accuracy for plant leaf disease classification.