Klasifikasi Penyakit Daun Tanaman Menggunakan MobileNetV4 Berbasis Transfer Learning: 116 - 131

Syifa Maulida (Institut Teknologi Kalimantan)
Tito Ariffianto Miftahul Huda (Institut Teknologi Kalimantan)
Wafiq Ajriyah (Institut Teknologi Kalimantan)
Rizky Amelia (Institut Teknologi Kalimantan)
Gusti Ahmad Fanshuri Alfarisy (Institut Teknologi Kalimantan)

Specta Journal of Technology
Vol. 10 No. 2 (2026): Specta Journal of Technology

Image-based plant leaf disease classification has become an important focus in agriculture to reduce significant losses and improve crop quality. This study proposes the use of Convolutional Neural Network (CNN) with MobileNetV4 architecture for plant leaf disease classification using transfer learning. MobileNetV4 was chosen because of its efficient ability to recognize visual objects with a light number of parameters, making it suitable for implementation on devices with limited resources. The dataset used is PlantVillage, which contains 20,600 labeled images across 15 disease categories. By using MobileNetV4 pre-trained on ImageNet, and modifying the final layers, we achieved a classification accuracy of 99.47%. This approach shows high potential for practical implementation on mobile devices due to its computational efficiency.