Autonomous Wildlife monitoring drone
Built by DARNISH S
Wildlife Sentinel-Aero: Autonomous Wildlife Monitoring Drone Wildlife Sentinel-Aero is an AI-powered autonomous drone for real-time wildlife detection, tracking, and population monitoring. The system combines autonomous flight with computer vision to monitor forests, wildlife sanctuaries, and agricultural boundaries without continuous human supervision. The drone follows predefined waypoints using GPS and a Pixhawk flight controller while capturing live aerial footage. YOLOv8s-Seg performs instance segmentation to accurately detect animals using polygon masks instead of bounding boxes. ByteTrack persistent tracking assigns a unique ID to each animal, ensuring individuals are counted only once and eliminating double-counting across frames. Each detection is tagged with GPS coordinates, altitude, and timestamps, enabling wildlife mapping and movement analysis. When animals or unauthorized humans are detected, the system automatically captures annotated images and sends instant Telegram alerts with location details. A browser-based dashboard provides live video streaming, real-time species counts, detection logs, telemetry data, and wildlife activity heatmaps. Recorded drone footage can also be analyzed to generate species-wise population reports and annotated output videos. Designed for remote deployment, Wildlife Sentinel-Aero operates fully offline and supports long-duration monitoring missions, making it an effective tool for conservation, biodiversity assessment, anti-poaching surveillance, and human-wildlife conflict prevention. Key Features * Autonomous drone navigation * YOLOv8s-Seg instance segmentation * ByteTrack unique animal counting * GPS-tagged detections * Real-time dashboard * Telegram alerts * Wildlife heatmaps * Offline operation * Population reporting