Research on Airport Target Recognition Based on YOLOv5 Algorithm
Mingsong Cao, Zhihao Zhang, Pingping Zhang · 2024
Traditional radar, infrared and other active electromagnetic reconnaissance means will radiate electromagnetic waves, which are easy to be detected and reacted by the enemy. The UAV carries visual reconnaissance equipment to fly over and near the enemy target for close in reconnaissance, which can covertly grasp the target and surrounding intelligence information. The vision equipment carried by each UAV sweeps the target area during flight, and identifies the acquired image in real time through the airborne computer and recognition algorithm to determine the type, number and distribution of targets in the reconnaissance area. This paper expounds the application of deep learning, neural network and other technologies in computer vision, improves the YOLOv5 algorithm and constructs the target data set to recognize and train the target on the airport. Through the model test, the accurate recognition of the selected aircraft target is realized, which provides a technical path for the key link of target recognition in UAV reconnaissance.