Ferdiyanto, Novriyan and Sutedi, Sutedi (2025) APLIKASI DETEKSI PENYAKIT TANAMAN JAGUNG BERBASIS ANDROID MENGGUNAKAN ALGORITMA CNN DAN MODEL TERLATIH TENSORFLOW LITE. Skripsi thesis, Institut Informatika dan Bisnis Darmajaya.
![[img]](http://repo.darmajaya.ac.id/style/images/fileicons/text.png) |
Text
1. Cover.pdf Download (199kB) |
|
Text
2. Halaman Pernyataan.pdf Download (312kB) |
|
Text
3. Halaman Persetujuan.pdf Download (532kB) |
|
Text
4. Halaman Pengesahan.pdf Download (607kB) |
|
Text
5. Riwayat Hidup.pdf Download (152kB) |
|
Text
6. Halaman Persembahan.pdf Download (148kB) |
|
Text
7. Halaman Motto.pdf Download (197kB) |
|
Text
8. Intisari.pdf Download (142kB) |
|
Text
9. Abstract.pdf Download (663kB) |
|
Text
10. Prakata.pdf Download (686kB) |
|
Text
11. Daftar Isi.pdf Download (256kB) |
|
Text
12. Daftar Tabel.pdf Download (198kB) |
|
Text
13. Daftar Gambar.pdf Download (215kB) |
|
Text
14. Bab 1.pdf Download (218kB) |
|
Text
15. Bab 2.pdf Download (466kB) |
|
Text
16. Bab 3.pdf Download (1MB) |
|
Text
17. Bab 4.pdf Download (3MB) |
|
Text
18. Bab 5.pdf Download (201kB) |
|
Text
19. Daftar Pustaka.pdf Download (217kB) |
|
Text
20. Lampiran.pdf Download (249kB) |
Abstract
Corn serves as a strategic commodity in Indonesia's agricultural sector, playing a crucial role as a source of food, animal feed, and industrial raw materials. However, corn production in several regions, such as Lampung, experienced a decline due to plant diseases. These diseases could be identified through symptoms visible on the leaves, such as leaf blight, common rust, and gray leaf spot. To assist farmers in detecting these diseases quickly and accurately, this study developed an Android-based application using a Convolutional Neural Network (CNN) algorithm with a pre-trained TensorFlow Lite model. The dataset consisted of four categories with 4,000 images—1,000 images per category: Healthy, Blight, Common Rust, and Gray Leaf Spot. The CNN model was evaluated using metrics such as accuracy, precision, recall, and F1- score, yielding an accuracy of 94.82% for the 70:30 training-testing ratio and 95.16% for the 80:20 ratio. The application enabled farmers to identify corn diseases simply by capturing an image, allowing faster and more effective response measures.
| Item Type: | Thesis (Skripsi) |
|---|---|
| Subjects: | Ilmu Komputer eSkripsi |
| Divisions: | Skripsi/TA & PKPM/KP - Fakultas Ilmu Komputer > Sistem Informasi |
| Depositing User: | Novriyan Ferdiyanto |
| Date Deposited: | 27 Aug 2025 08:00 |
| Last Modified: | 27 Aug 2025 08:00 |
| URI: | http://repo.darmajaya.ac.id/id/eprint/21061 |
Actions (login required)
 |
View Item |