Analisis dan Implementasi Diagnosis Penyakit Sawit dengan Metode Convolutional Neural Network (CNN)
Jambi Province is a producer of palm oil as a mainstay of commodities. However, the limited insight of farmers in Jambi to oil palm pests and diseases affects oil palm productivity. Meanwhile, knowing the types of pests and diseases in oil palm requires an expert, but access restrictions are a problem. This study offers a diagnosis of oil palm disease using the most popular concept in the field of artificial intelligence today. This method is deep learning. Various recent studies using CNN, say the results of image recognition accuracy are very good. The data used in this study came from oil palm image data from the Jambi Provincial Plantation Office. After the oil palm disease image data is trained, the training data model will be stored for the process of testing the oil palm disease diagnosis. The test evaluation is stored as a configuration matrix. So that it can be assessed how successful the system is to diagnose diseases in oil palm plants. From the testing, there were 2490 images of oil palm labeled with 11 disease categories. The highest accuracy results were 0.89 and the lowest was 0.83, and the average accuracy was 0.87. This shows that the results of the classification of oil palm images with CNN are quite good. These results can indicate the development of an automatic and mobile oil palm disease classification system to help farmers.