Comparison of Different Classifiers for Automatic Target Recognition Systems

Within unmanned aerial vehicles on-going research topics, automatic target recognition is acquiring relevance. This is due to the easiness with which it is possible to acquire such devices. To do so, signal processing and classification techniques could be used. Many signal processing techniques can be used, and the literature usually rely on the use of time-frequency distributions. But, since there is a high computational cost involved, the authors worry about the use of such mathematical techniques, and a low computationally expensive technique is applied to create an automatic target recognition system. To accomplish this, a monostatic K-band radar system is used to send and receive continuous electromagnetic signals, which are processed with fast Fourier transform for feature extraction to, finally, be applied on an artificial neural network and a support vector machine approach. The classification results between both techniques are analyzed and the best solution is selected.