RAMADHAN, FAJRI (2026) PENGEMBANGAN SISTEM KLASIFIKASI PENYAKIT DAUN PADI MENGGUNAKAN PENDEKATAN HYBRID EFFICIENTNETV2-B3 DAN MOBILENETV2. S1 thesis, Universitas Mercu Buana Jakarta.
|
Text (HAL COVER)
Cover.pdf Download (613kB) | Preview |
|
![[img]](https://repository.mercubuana.ac.id/style/images/fileicons/text.png) |
Text (BAB I)
Bab 1.pdf Restricted to Registered users only Download (158kB) |
|
|
Text (BAB II)
Bab 2.pdf Restricted to Registered users only Download (607kB) |
|
|
Text (BAB III)
Bab 3.pdf Restricted to Registered users only Download (847kB) |
|
|
Text (BAB IV)
Bab 4.pdf Restricted to Registered users only Download (1MB) |
|
|
Text (BAB V)
Bab 5.pdf Restricted to Registered users only Download (98kB) |
|
|
Text (DAFTAR PUSTAKA)
Daftar Pustaka.pdf Restricted to Registered users only Download (163kB) |
|
|
Text (LAMPIRAN)
Lampiran.pdf Restricted to Registered users only Download (627kB) |
Abstract
Rice (Oryza sativa) is one of the main staple food commodities in Indonesia, whose productivity is often reduced due to attacks by leaf diseases such as Bacterial Leaf Blight, Brown Spot, Leaf Blast, and several other types of diseases. Delays in the disease identification process can accelerate the spread of infection and negatively impact crop yields. This study aims to develop a rice leaf disease classification system using a hybrid Convolutional Neural Network (CNN) approach combining EfficientNetV2-B3 and MobileNetV2 architectures. The dataset used is the Rice Leaf Diseases Detection Dataset, consisting of 8,000 images across 8 rice leaf disease classes. Before the training process, the images undergo a segmentation stage using HSV and Otsu Thresholding methods. This is followed by preprocessing, image resizing, and data augmentation to increase dataset variability. The model is built using pretrained ImageNet weights and trained using the Adam optimizer, with callbacks applied to maintain training stability. The results show that the developed model achieves a validation accuracy of 98.75% and a testing accuracy of 98.42%.. Keywords: Convolutional Neural Network, EfficientNet, MobileNet, Rice Leaf Disease Detection. Padi (Oryza sativa) merupakan salah satu komoditas pangan utama di Indonesia yang produktivitasnya sering menurun akibat serangan penyakit daun seperti Bacterial Leaf Blight, Brown Spot, Leaf Blast, dan beberapa jenis penyakit lainnya. Keterlambatan dalam proses identifikasi penyakit dapat mempercepat penyebaran infeksi serta berdampak pada penurunan hasil panen. Penelitian ini bertujuan untuk mengembangkan sistem klasifikasi penyakit daun padi menggunakan pendekatan hybrid Convolutional Neural Network (CNN) dengan kombinasi arsitektur EfficientNetV2-B3 dan MobileNetV2. Dataset yang digunakan berasal dari Rice Leaf Diseases Detection Dataset dengan total 8000 citra yang terdiri dari 8 kelas penyakit daun padi. Sebelum proses pelatihan, citra terlebih dahulu melewati tahap segmentasi menggunakan metode HSV dan Otsu Thresholding. Selanjutnya dilakukan preprocessing, resizing gambar, serta augmentasi data untuk meningkatkan variasi dataset. Model dibangun dengan memanfaatkan bobot pretrained ImageNet dan dilatih menggunakan optimizer Adam, serta dilengkapi dengan callback untuk menjaga kestabilan proses training. Hasil penelitian menunjukkan bahwa model yang dikembangkan mampu mencapai validation accuracy sebesar 98,75% dan akurasi pengujian sebesar 98,42%. Kata Kunci : Convolutional Neural Network, EfficientNet, MobileNet, Deteksi Penyakit Padi.
Actions (login required)
 |
View Item |
