A Low-Slow-Small UAV Radar Target Recognition Method Based on Point-Track Feature Analysis and Multifeature Fusion

Unmanned aerial vehicle (UAV) targets are characterized by low-altitude flight, small velocity, and small radar cross section (RCS), making them prone to being confused with small low-altitude flying targets such as birds in radar target recognition scenarios. To address these challenges, this article proposes a radar target recognition method integrating kinematic characteristic classification and multifeature fusion (MFF). The approach strategically partitions recognition targets into distinct category subspaces through kinematic analysis. We have put forward a novel 'point-track-angle' (PTA) feature, which serves as an effective tool for clutter recognition. In the recognition of low-slow-small (LSS) targets, we utilize different classifiers to recognize the PTA features, RCS, and micro-Doppler features, and then conduct further decision-level fusion of the classification results. Field experiments demonstrate significant performance improvement in distinguishing low-altitude small UAVs from avian targets. The recognition accuracy for both UAV and bird target points exceeds 92%, and the track recognition accuracy based on sequential statistics exceeds 95%. The target features utilized by this method are relatively easy to obtain, and it imposes low requirements on radar system architecture and resolution. It has been successfully deployed on the UAV surveillance radar developed by our laboratory.