WildEye: An AI-IoT Smart Forest System for Fire Detection and Wildlife-Vehicle Collision Prevention

There has been a decline in wildlife protection due to reasons like deforestation, urbanization, changes in climate, and other human activities. These changes have led to serious issues such as animal human-collision, forest fires, loss of biodiversity and pollution. The Wildeye project is the proposed solution to handle all these challenges. This project integrates both Iot and Artificial Intelligence for continuous forest observation. The system is integrated with smart web cameras and automatic braking system which helps in tracking the animals and the system automatically applies brake to avoid animal- vehicle collision. This setup uses Yolo5 algorithm which is a real time object detection algorithm to detect and track animals that helps in preventing wildlife-vehicle collision. In certain cases when the automatic braking system fails and the collision takes place, the system sends an immediate alert to the forest department for the further rescue of the animal. The approach also makes use of smoke and fire sensors to detect and prevent forest fires and sends an immediate alerts and location of the incident to the responsible authorities. The system shows improvement of Mean Average Precision at IoU with 15%, Precision of 6.2% while Frames per Second (FPS) 17.6% in comparison with Yolo2 and shows much better results than CNN algorithms.