Animal Intrusion Detection System using YOLOv11

Crop damage by wildlife, particularly monkeys, pigs, and squirrels, poses a significant challenge to farmers in Tamil Nadu, India. This paper presents a real-time animal detection and notification system designed to mitigate this problem. Utilizing YOLOv11 with a custom-trained dataset, the system detects these animals in real time using CCTV IP camera feeds. Upon detection, the system triggers immediate alerts to farmers via VoIP calls and messages, enabling prompt intervention. Real-time field monitoring is also provided through a database interface. This cost-effective and accessible solution empowers farmers to protect their crops and improve agricultural productivity. Field tests demonstrated a mean Average Precision (mAP) of 0.84 for monkey detection, validating the system's accuracy. Real-time processing at 20 frames per second enables timely alerts, promising a significant reduction in crop damage and improved farmer livelihoods. Future development will focus on automated deterrent systems and data analysis for predictive modeling and informed decision-making.