Robby, Mandala (2026) IMPLEMENTASI DEEP LEARNING DETEKSI OTOMATIS PENYAKIT PADA DAUN PADI DENGAN METODE YOLOv11. Diploma thesis, UIN RADEN INTAN LAMPUNG.
|
PDF
Restricted to Repository staff only Download (3MB) |
|
|
PDF
Restricted to Repository staff only Download (4MB) |
Abstract
ABSTRAK Penyakit tanaman padi seperti bacterial leaf blight, leaf blast, dan brown spot merupakan ancaman serius yang dapat menurunkan produktivitas pangan nasional secara signifikan. Identifikasi penyakit secara visual oleh petani masih tidak efisien dan rentan menghasilkan diagnosis yang kurang tepat. Penelitian ini bertujuan untuk mengimplementasikan teknologi deep learning dengan algoritma YOLOv11 untuk mengembangkan sistem deteksi otomatis penyakit pada daun padi. Metode penelitian yang digunakan adalah Research and Development (R&D) yang mencakup tahapan pengumpulan dataset, pelabelan, preprocessing, pelatihan model, hingga implementasi ke dalam aplikasi web berbasis HTML, CSS, dan JavaScript menggunakan TensorFlow.js. Dataset yang digunakan terdiri dari 3.087 citra hasil augmentasi yang dibagi menjadi data latih (70%), validasi (20%), dan uji (10%). Model dilatih selama 100 epoch di lingkungan Google Colab. Hasil pengujian pada data uji menunjukkan bahwa model mencapai nilai Mean Average Precision ([email protected]) rata-rata sebesar 86.8%. Model menunjukkan performa sangat tinggi dalam mendeteksi daun sehat (mAP 97.6%), namun akurasinya lebih rendah untuk kelas penyakit spesifik. Sistem berbasis web yang dihasilkan berhasil mendeteksi penyakit secara fungsional. Penelitian ini menyimpulkan bahwa implementasi YOLOv11 efektif secara teknis, namun akurasi model untuk diagnosis penyakit spesifik masih memerlukan peningkatan lebih lanjut sebelum dapat diandalkan sepenuhnya di lapangan. Kata Kunci: Deep Learning, Deteksi Penyakit Padi, Pengolahan Citra, Sistem Deteksi Otomatis, YOLOv11. ABSTRACT Rice plant diseases, such as bacterial leaf blight, leaf blast, and brown spot, pose a serious threat to national food productivity. Manual visual identification by farmers remains inefficient and prone to misdiagnosis. This study aims to implement deep learning technology using the YOLOv11 algorithm to develop an automated detection system for rice leaf diseases. The Research and Development (R&D) methodology was employed, encompassing stages from dataset collection, labeling, and preprocessing to model training and implementation into a web-based application using HTML, CSS, JavaScript, and TensorFlow.js. The dataset consisted of 3,087 augmented images, partitioned into training (70%), validation (20%), and testing (10%) sets. The model was trained for 100 epochs within the Google Colab environment. Results from the testing phase indicated that the model achieved an average Mean Average Precision ([email protected]) of 86.8%. While the model demonstrated high performance in detecting healthy leaves (mAP 97.6%), the accuracy for specific disease classes was comparatively lower. The resulting web-based system is functionally capable of detecting diseases. This study concludes that while the implementation of YOLOv11 is technically effective, further refinement in diagnostic accuracy for specific diseases is required before the system can be fully reliable for field applications. Keywords: Deep Learning, Rice Disease Detection, Image Processing, Automated Detection System, YOLOv11.
| Item Type: | Thesis (Diploma) |
|---|---|
| Subjects: | Sistem Informasi |
| Divisions: | Fakultas Sains dan Teknologi > Sistem Informasi |
| Depositing User: | LAYANAN PERPUSTAKAAN UINRIL REFERENSI |
| Date Deposited: | 20 Jan 2026 07:46 |
| Last Modified: | 20 Jan 2026 07:46 |
| URI: | https://repository.radenintan.ac.id/id/eprint/42336 |
Actions (login required)
![]() |
View Item |


