Real-time detection of multi-scale miniature unmanned aerial vehicles based on YOLOv9

Abstract

Aiming at the security risks of unregulated and unmanned aerial vehicles (UAVs), this paper proposes a new real-time detection method based on YOLOv9, which integrates reversible functions, programmable gradient information, and a generalized high-efficiency layer aggregation network, and combined with downsampling and local feature training method. Experiments show that the detection accuracy of the method is more than 90% and the processing frame rate is more than 20Hz.