The research develops an autonomous UAV-based bird scaring system for agriculture. It includes bird flock detection using machine learning and image segmentation (K-means clustering), automatic height adjustment to avoid obstacles, and activation of an onboard vibrator-sound system to scare birds, enabling continuous crop monitoring and protection.