A novel approach for automatic detection and tracking of flying objects

Abstract

In this study, a new method is presented to automatically detect and track flying objects through video systems that are used for surface to air tracking tasks. In this approach, a method has been developed in which Standard Deviation is used to determine the presence of a flying object. The measurement data is adapted to track, so that the flying object becomes more dominant than the background. In order to track the detected target in real time, Interacting Multiple Model Probabilistic Data Association with Amplitude Information (IMMPDA-AI) algorithm is used. Although the IMMPDA-AI algorithm is mainly a point tracking algorithm, in this study, its applicability to video tracking is shown. For this purpose, the amplitude information of the sampled video frames is encoded as point data and the tracking is performed on this data. Thus, an algorithm has been developed in which the target is automatically detected, track initiated and continued. The algorithm is evaluated for different maneuvers, target types and clutter situations, and successful results are obtained.