Deteksi Dan Segmentasi Kerusakan Tanaman Padi Akibat Serangan Hama Menggunakan YOLOv8 dan U-Net Untuk Pemetaan Persebaran Kerusakan Berbasis Grid

Azhari, Taufiq Septiyawan Azhari (2026) Deteksi Dan Segmentasi Kerusakan Tanaman Padi Akibat Serangan Hama Menggunakan YOLOv8 dan U-Net Untuk Pemetaan Persebaran Kerusakan Berbasis Grid. Other thesis, Institut Teknologi Sepuluh Nopember.

Text 2040221004-Undergraduate_Thesis.pdf - Accepted Version Restricted to Repository staff only Download (6MB) | Request a copy

Abstract

Padi merupakan komoditas pertanian yang berperan penting dalam mendukung ketahanan pangan nasional. Di Kabupaten Bojonegoro, tanaman padi rentan mengalami kerusakan akibat serangan hama wereng dan penggerek batang (sundep). Proses identifikasi kerusakan tanaman yang masih dilakukan secara manual memerlukan waktu dan bergantung pada pengamatan di lapangan, sehingga diperlukan suatu sistem yang dapat membantu proses identifikasi dan pemantauan kondisi tanaman. Penelitian ini bertujuan mengembangkan sistem deteksi dan segmentasi kerusakan tanaman padi akibat serangan hama menggunakan metode YOLOv8 dan U-Net berbasis citra drone. YOLOv8 digunakan untuk mendeteksi jenis kerusakan tanaman akibat serangan hama wereng dan sundep, sedangkan U-Net digunakan untuk melakukan segmentasi area kerusakan sehingga dapat dihitung persentase kerusakan tanaman. Hasil deteksi dan segmentasi kemudian dipadukan dengan data telemetry drone untuk menghasilkan pemetaan persebaran kerusakan berbasis grid secara non-realtime. Berdasarkan hasil pengujian, model YOLOv8 memperoleh tingkat prediksi benar sebesar 75% pada kelas Sundep dan 66% pada kelas Wereng. Model U-Net menghasilkan nilai accuracy sebesar 94%, precision 76%, recall 52%, IoU 45%, dan Dice coefficient 62%. Sistem juga berhasil menampilkan hasil deteksi, segmentasi, persentase kerusakan, serta pemetaan persebaran kerusakan tanaman melalui dashboard. Hasil penelitian menunjukkan bahwa kombinasi YOLOv8 dan U-Net dapat digunakan untuk mendeteksi jenis kerusakan tanaman padi, melakukan segmentasi area kerusakan, serta menyajikan informasi persebaran kerusakan berbasis grid menggunakan citra drone.
==================================================================================================================================
Rice is an important agricultural commodity that supports national food security. In Bojonegoro Regency, rice crops are susceptible to damage caused by brown planthopper and rice stem borer (sundep) infestations. Manual identification of crop damage requires considerable time and depends on field observations, creating the need for a system that can assist in identifying and monitoring crop conditions. This study aims to develop a drone image-based system for detecting and segmenting rice plant damage caused by pest attacks using YOLOv8 and U-Net. YOLOv8 is employed to detect the type of rice plant damage caused by brown planthoppers and stem borers, while U-Net is used to segment the damaged areas and estimate the percentage of damage. The detection and segmentation results are integrated with drone telemetry data to generate a non-real-time grid-based damage distribution map. Experimental results show that the YOLOv8 model achieved correct prediction rates of 75% for the Sundep class and 66% for the Wereng class. The U-Net model achieved an accuracy of 94%, precision of 76%, recall of 52%, Intersection over Union (IoU) of 45%, and a Dice coefficient of 62%. The developed system successfully visualizes detection results, segmented damage areas, damage percentages, and grid-based damage distribution through an integrated dashboard. These results indicate that the combination of YOLOv8 and U-Net can be used to detect rice plant damage types, segment damaged areas, and provide grid-based damage distribution information from drone imagery.

Item Type:	Thesis (Other)
Uncontrolled Keywords:	Deteksi Kerusakan Tanaman Padi, Segmentasi Citra, YOLOv8, U-Net, Pemetaan Berbasis Grid. Rice Plant Damage Detection, Image Segmentation, YOLOv8, U-Net, Grid-based Mapping.
Subjects:	S Agriculture > SB Plant culture > SB191.R5 Rice farming T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK5105.546 Computer algorithms T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK6592.A9 Automatic tracking. T Technology > TK Electrical engineering. Electronics Nuclear engineering > TK7888.3 Digital computers
Divisions:	Faculty of Vocational > 36304-Automation Electronic Engineering
Depositing User:	Taufiq Septiyawan Azhari
Date Deposited:	05 Aug 2026 04:11
Last Modified:	05 Aug 2026 04:11
URI:	http://repository.its.ac.id/id/eprint/143998

Actions (login required)

View Item