M. Irwan, Khairun Saddami, Rizka Ramadhana, Ramzi Adriman
Rice field rats (Rattus argentiventer) are the primary cause of post-harvest rice yield losses in Indonesia, with potential losses reaching up to 37% per hectare and a decline in grain quality. Traditional control methods, such as traps and rodenticides, are often ineffective on a large scale and have negative environmental impacts. This study developed a real-time rat detection system based on deep learning using the YOLOv11 algorithm, implemented on a Raspberry Pi 5 for early monitoring in rice storage warehouses. A total of 410 rat images were collected and processed through Roboflow for segmentation, then divided into training (328), validation (41), and testing (41) datasets. The YOLOv11 model was trained for 500 epochs with a learning rate of 0.002, batch size of 4, and the AdamW optimizer. To prevent overfitting, early stopping was applied with a threshold of 100 epochs without performance improvement. Training results showed the model achieved a precision of 0.955, recall of 0.86, mAP@0.50 of 0.87, and mAP@0.50-0.95 of 0.39. The prototype system, consisting of a Raspberry Pi 5, webcam, and buzzer, was capable of real-time rat detection and provided audible alerts. This research demonstrates that integrating YOLOv11 with low-power computing devices offers an effective, sustainable, and applicable solution for rodent pest detection in agricultural storage environments. © 2025 IEEE.
Universitas Syiah Kuala, Banda Aceh, Indonesia; Universitas Syiah Kuala, Departement of Electrical and Computer Engineering, Banda Aceh, Indonesia
Research at a Glance
Register to unlock
Topics & SDG Alignment
Register to unlock
Collaboration
Register to unlock
Author Profile (Selected)
Register to unlock
References Overview
Register to unlock
Journal & Source
Register to unlock
Metadata & Integrity
Register to unlock