Radar Intelligent Detection Approach for Swarm Target Scene

Unmanned aerial vehicle (UAV) swarms have been widely used in civil and military fields. UAV swarm has the characteristics of multi-scale irregularity in target distribution and small individual targets. This brings certain challenges to the traditional radar target detection algorithms and deep learning algorithms. Therefore, this paper proposes a novel point target detection approach for UAV swarm scene. For the multi-scale and irregular characteristics of target distribution within the UAV swarm, the detection box of each overall swarm target can be obtained by combining the multi-scale superposition and the feature pyramid (FPN) network structure. Subsequently, the thresholds of the detection cells inside the boxes are estimated by the background noise outside the detection boxes, so as to realize the detection of the point targets inside each swarm. Simulation results show that the proposed approach outperforms YOLOv7 and several classical constant false alarm rate(CFAR) detection algorithms in the radar field.