Aerial Target Identification Based on Classification Learner

Determination of radar traces are world-widely important issue for the defense industry and civil aviation. Due to many objects such as drones, birds, higher energy plants components, there have been many artifacts to determine real flying object (i.e. target) from these surrounding factors mentioned. Therefore, a series of target classes are of critical importance in surveillance radars, which are used to manage air traffic or detect targets within the territory. In the present work, we report a comprehensive comparison of various machine learning algorithms by considering the flight data series of the objects. Different machine learning algorithms are applied to the problem by considering the statistical features the obtained 3 typical flight data, namely, real and synthetically created S band Radar Cross Section (RCS), Velocity and Altitude values. Following the analyses, it has been proven that the proposed model can diagnose the classes of air targets with good accuracies namely from 85.71% to 90.16% during the testing phase of 5 different machine learning algorithms.