Aircraft Target Recognition: A novel approach for features extraction from ISAR images

In this paper, we present a system for Automatic Target Recognition (ATR) based on ISAR images. The methodology used is based on knowledge discovery from data (KDD process) process adapted to radar field. The shape extraction is the most important step in recognition system. However, we propose a new approach for Target shape extraction based on combination of Smallest Univalue Segment Assimilating Nucleus (SUSAN) method and Variational Level Set (VLS). The feature vector is then represented by Fourier descriptors of each target shape. Finally, recognition scheme is achieved by both: Support Vectors Machine (SVM) and K Nearest Neighbors (KNN) classifiers.