Anti-Occlusion UAV Tracking Algorithm with a Low-Altitude Complex Background by Integrating Attention Mechanism

In recent years, the increasing number of unmanned aerial vehicles (UAVs) in low-altitude airspace have not only brought convenience to people’s work and life, but also great threats challenges. process UAV detection tracking, there are common problems such as target deformation, occlusion, targets being submerged by complex background clutter. This paper proposes an anti-occlusion tracking algorithm for backgrounds integrating attention mechanism that mainly solves occlusion when UAVs. First, extracted features enhanced using SeNet mechanism. Second, occlusion-sensing module is used judge whether occluded. If occluded, continues. Otherwise, LSTM trajectory prediction network predict position subsequent frames flight before occlusion. study was verified on OTB-100, GOT-10k integrated datasets. The accuracy success rate datasets were 79% 50.5% respectively, which 10.6% 4.9% higher than those SiamCAM algorithm. Experimental results show could robustly track a small background.