Revitalizing Strawberry Leaves: Developing a Tipburn and Leaf Spot Disease Detection System Through Convolution Analysis Using CNN Method

Sari, Erika Lety Istikhomah Puspita

JIM: Jurnal Ilmiah Mahasiswa Pendidikan Sejarah
Vol 8, No 3 (2023): Juni, socio-economics, community law, cultural history and social issues

Currently, plant identification processes are still manual and fraught with difficulties due to human nature. Human error can render the desired results ineffective. Another issue is that diseases in strawberry plants, such as tipburn and leaf spot, can hinder growth, affect plant quality, and have economic implications for agriculture. Therefore, the researchers developed a deep learning model using the Convolutional Neural Network (CNN) algorithm, specifically VGG16, with a dataset of 2,897 photos. The aim was to classify tipburn, leaf spot, and healthy states of strawberry plant leaves. To minimize overfitting during the classification training, the training dataset was included. This was done to enable the model to recognize fundamental variations of strawberry leaf images and achieve training and validation accuracies of 95.05% and 97.4%, respectively. Consequently, the training loss value was 19.68%, while the validation loss value was only 7.54%. The findings showed accuracies greater than 90% for both training and validation parameters. This research is expected to be beneficial in providing information on data augmentation processes and disease classification in strawberry plants.