Klasifikasi Penyakit pada Buah Jeruk Berdasarkan Citra dengan Pendekatan Transfer Learning Menggunakan Arsitektur Densenet-121

Sheli Agustina (Universitas Bina Insan)
Asep Toyib Hidayat (Universitas Bina Insan)
Satrianansyah (Universitas Bina Insan)
Rudi Kurniawan (Universitas Bina Insan)

Scientific Journal of Informatics
Vol. 10 No. 1 (2025): Jurnal Ilmiah Informatika

This study aims to develop a classification system for citrus fruit diseases based on digital images using a machine learning approach. The primary challenge in citrus cultivation is disease attacks that affect both the quality and quantity of production. In this research, image processing techniques were applied to extract color, shape, and texture features from citrus fruit images, which were then used as input for classification algorithms. This study uses the DenseNet-121 architecture for orange fruit image classification. The dataset used consisted of images of healthy citrus fruits and those affected by various diseases, such as blackspot, canker, and greening. The testing results showed that the DenseNet-121 architecture achieved the highest accuracy in classifying citrus diseases, with an accuracy rate of up to 99%. This system is expected to assist farmers and relevant stakeholders in early disease detection and in taking appropriate control measures.