<![CDATA[Abstract Indonesia, as an agricultural country, has a very vital agricultural sector, including the cultivation of cayenne pepper. Cayenne peppers are often infected with anthracnose disease caused by the fungus Colletotrichum sp., causing significant economic losses. This research aims to develop a Convolutional Neural Network (CNN) model for classifying anthracnose in images of cayenne pepper, in order to increase the effectiveness of disease diagnosis. Image data was obtained from chili gardens in Savanajaya Village, Buru Regency, with a total of 1000 images, which were divided into 500 images of healthy chilies and 500 images of infected chilies. The data is processed and labeled manually, then resized for consistency. CNN was trained using the Adam optimizer, RMSprop, and SGDM, with test results showing that the Adam optimizer provided the highest accuracy of 93.25%. The implementation of CNN has proven effective in classifying anthracnose, helping farmers in making timely decisions for disease control, thereby increasing productivity and reducing economic losses. This research emphasizes the importance of choosing the right optimizer and dataset quality in developing image-based plant disease classification models.]]>
Copyrights © 2024
