Detection of Animals on Highways During Night

The detection of animals on highways during night-time is crucial for enhancing road safety and reducing accidents involving wildlife. This project focuses on leveraging the YOLOv12 object detection model to identify various animal species on highways in real-time, using nighttime imagery. The model is trained to detect several animal classes, including birds, boars, cats, deer, dogs, opossums, persons, raccoons, skunks, and squirrels, achieving high performance across various evaluation metrics such as precision, recall, and F1 score. The model's efficiency is evaluated based on its ability to correctly classify these animals in challenging nighttime conditions. The system utilizes the YOLOv12 architecture for its speed and accuracy in detecting animals from images captured by highway cameras. By integrating this model into traffic monitoring systems, it aims to provide real-time alerts for potential animal-related road hazards, thus contributing to road safety efforts. The system is trained and evaluated using Python, with libraries such as OpenCV for image processing and TensorFlow for model implementation. This approach provides a scalable and efficient solution for mitigating the risks associated with wildlife crossings on highways.

Keywords: YOLOv12, Animal Detection, Road Safety, Object Detection, Traffic Monitoring, Real-Time Alerts, Nighttime Imaging, Machine Learning, Deep Learning, Image Classification, Highway Safety, Wildlife.