Research on air-to-air UAV tracking method based on deep learning
Hexiang Hao, Yueping Peng, Baixuan Han, Wenchao Liu · 2024
With the rapid development of UAV technology and group intelligence technology, hundreds of small-scale, low-cost UAVs in the form of self-organized networks to perform complex combat tasks, or in the form of a single frame to perform reconnaissance, raid tasks, to the low-altitude airspace defense has brought great threats and challenges. Based on the security needs of low-altitude airspace defense, the first task to implement countermeasures against enemy UAVs is to detect and track the incoming UAVs using visual detection technology. This paper firstly introduces the current situation in the field of air-to-air UAV tracking based on UAV perspective, combs through the classical and the latest deep learning-based target tracking algorithms, and analyzes and compares between the algorithms; secondly, it introduces the tracking datasets and evaluation indexes applicable to this task; finally, it gives an outlook of the future direction of the development of the UAV tracking task based on UAV perspective in the air-to-air UAV tracking.