Human–wildlife conflict in agricultural areas leads to significant crop losses and livestock threats, creating an urgent need for reliable real-time detection systems. This study presents the Wildlife Intrusion Detection System (WAIDS), a novel IoT-enabled solution designed to mitigate such risks. The system integrates PIR motion sensors, a Raspberry Pi computing unit, and a custom-trained YOLOv8n object detection model for robust wildlife identification, with Twilio SMS alerts ensuring rapid farmer response. A strategically deployed sensor network captures activity along the farm perimeter, while the Raspberry Pi executes YOLOv8n inference for accurate classification. A dataset comprising diverse animal images under varying conditions (day/night, weather, and motion speeds) was curated for training and testing. The system achieved 80–85% detection accuracy, with evaluation metrics of precision (0.xx), recall (0.xx), F1-score (0.xx), mean Average Precision (mAP) (0.xx), and average inference latency of xx ms per frame. These results highlight the system’s robustness under real-world field conditions, making it suitable for practical deployment. The proposed WAIDS significantly enhances farm security, minimizes agricultural losses, and demonstrates the potential of IoT and deep learning integration for sustainable agriculture and wildlife management.
Patkar et al. (Sat,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: