In recent years, there has been a growing interest in automatic tracking systems. These systems have revolutionary potential in a wide variety of fields, including security, sports, tracking and navigation. The DeepSORT algorithm is one of the most promising tracking algorithms. Combining it with deep learning methods achieves extraordinary results. The algorithm makes use of a Kalman Filter, which evaluates the dynamics of objects in the scene, thus enabling tracking even in cases of occlusion and motion of objects. This makes the algorithm an ideal candidate for use in real-world applications. In this project we will use this approach to track aircraft, vehicles and people. YOLOv5s algorithm is used for object detection. In order to improve the results for aircraft, we will propose a method for features extracting for aircraft – Siamese network with a triplet loss function.