An Algorithm for Small Aircraft Recognition by Unmanned Air Vehicle Imagery

Abstract The development of technology has made it possible to use images captured by unmanned aerial vehicles (UAVs) to recognize objects on the Earth’s surface, including aircraft. The high resolution of UAV images allows the implementation of significantly simpler algorithms than those using deep learning and neural networks. The paper proposes an algorithm for small aircraft recognition based on matching with reference images of aircraft (templates). The algorithm consists of three stages. At the first stage pre-filtering of the images is carried out in order to achieve better distinction between aircraft and background. On the second stage a morphological image processing and image segmentation are applied. On the third stage a correlation processing is used to compare the detected segments potentially containing an aircraft with the templates. In all studied cases the aircraft were successfully recognized and there wasn’t any false recognition.