Anti-UAV Detection Method Based on Local-Global Feature Focusing Module

When applied to the task of small UAV target detection on visual images, YOLOv5 algorithm suffered false positive and false negative a lot. An improved method based on YOLOv5 is proposed to deal with this situation. A Local-Global Feature Focusing Module is designed at the input of the backbone network. By increasing the context between local features and global features in the shallow layer, strengthened feature representation ability, and improved the detection effect of the drone target. Experiments on the test set show that compared with the method before the improvement, the proposed method's precision is increased by 1.41% and recall is increased by 2.96%. At the same time of ensuring the real-time detection, the effect on UAV targets detection of the proposed method is significantly increased.