Plant disease is a challenge in the agricultural sector, especially for rice farmers. Identification of diseases on rice leaves is the first step to eradicating and treating diseases, to minimize crop failure. With the rapid development of the convolutional neural network (CNN), rice leaf disease can be recognized well without the help of an expert. The MobileNet-V2 architecture is used to classify rice leaf diseases due to its small size but good performance. To improve the performance of the CNN model, a hyperparameter consisting of an epoch, batch size, learning rate, and optimizer. This study purpose to have hyperparameters optimal The dataset used consists of 3 classes of diseases that attack the leaves of rice plants, including blast, blight, and tungro. Based on the experiments that have been carried out, the determination of hyperparameters greatly influences the model performance. Hyperparameter with epochs, batch sizes 32 learning rate and optimizer gives the most optimal results with accuracy 97.56%, precision 97.64%, recall 97.57%, and f1-score 97.57%.
Copyrights © 2022