<![CDATA[Indonesia has rich biodiversity, including various types of herbal plants commonly used in traditional medicine. However, manual identification and classification of herbal leaves remain challenging due to similarities in shape, color, and texture between species. This study aims to develop an automatic classification system for herbal leaf types using the Convolutional Neural Network (CNN) method with the AlexNet architecture. The dataset used in this research was obtained from a public Kaggle repository and consists of ten classes of herbal leaves, namely guava, curry, basil, turmeric, mint, papaya, betel, soursop, aloe vera, and green tea. Image preprocessing includes resizing to 224×224 pixels, normalization, and data augmentation to improve model generalization. The model was trained and tested using an 80:20 data split with several experimental configurations. The best performance was achieved at a learning rate of 0.0001, batch size of 32, and 25 epochs, resulting in a training accuracy of 88.25% and a testing accuracy of 73.50%. These results show that AlexNet can effectively extract visual features from herbal leaf images and perform accurate classification. This study demonstrates that CNN-based classification is an efficient approach for recognizing herbal plants automatically.]]>
Copyrights © 2025
