Two different novel methods for classification of aircraft categories of Inverse Synthetic Aperture Radar (ISAR) images are presented. The first method forms numerical equivalents to shape, size, and other aircraft features as critical criteria to constitute the algorithm for their correct classification. The second method compares each ISAR image to unions of images of the different aircraft categories. ISAR images are constructed based on the Doppler shifts of various parts, caused by the rotation of the aircraft and the radar reflection pulse shape, which includes the size or duration of the radar pulse. The proposed classification algorithms were tested on seven aircraft categories. All seven different aircraft models are flying a holding pattern. The aim of both algorithms is to quickly match and determine the similarity of the captured aircraft to the seven different categories where the aircraft is in any position of a prescribed holding pattern. Experimental results clearly indicate that in most parts of the holding pattern the category of the aircraft can be successfully identified with both proposed methods. The union method shows more successful identification results and is superior to the results we obtained in the first proposed method.
Top1. Introduction
Inverse Synthetic Aperture Radar (ISAR) acquires both range and Doppler shifts to form an image of a target. It offers more information about the target than a one dimensional range profile obtained from timing and duration of a radar echo return alone. Any rotation of the target introduces Doppler shifts in frequency from different parts of the target. Together with pulse return shape, these Doppler shifts reveal information about the shape and the size of the target. Some methods test only for features that are extracted from the ISAR image to identify what kind of target has been detected, as described in (Saidi, 2008), (Maskall, 2002), (Vespe, 2006), (Manikandan, 2007). This is more efficient than evaluating the whole target. Some methods use optimal classifiers, presented in (Kim, 2005), (Martorella, 2008), (Hu, 2007), to determine what kind of target is responsible for the image. In military applications, it is desirable to be able to automatically identify the aircraft model from as little radar information as possible. It is likely that the combat crews of attacking aircraft will do everything they can to deprive their target of the use of radar information about their attack. Usually, automatic electronic countermeasure avionics on the attacking aircraft or an electronic warfare officer in the flight crew will jam the radar as soon as the radar is detected. The target of the attack will have only the first radar pulses to try to identify what categories of aircraft are making the attack. It would be a great advantage to know from as little as one returned radar pulse what kind of aircraft are involved.
Many schemes of ISAR target recognition have been developed that test only the features taken from the ISAR images with high efficiency. This saves the time exhausted while trying to compare the whole target aircraft. Some methods use optimal classifiers to determine what kind of target is responsible for the image.
Algorithms for automatic classification of various aircraft models are proposed in this chapter. Just as in direct human observation of a target such as an aircraft there are limits of what an observer, in this case the proposed automatic software can see and this will make it difficult to identify what classification of aircraft is observed. There are situations in which there are limits of what is possible to determine from an ISAR image. These limits will be discussed and possible ways to get more information out of the ISAR images will be suggested.