Robust 3D ISAR Ship Classification

Classification of inverse synthetic aperture radar (ISAR) ship imagery is difficult due to the motion of the sea causing a wide variation of the observed images. The three dimensional (3D)-ISAR technique was developed as an alternative representation with the target represented by a 3D point cloud. In this paper, we compare two different approaches to classification using 3D-ISAR point clouds of maritime targets. The first approach makes use of features extracted from the 3D-ISAR generated point cloud of the target from different perspectives (i.e. side, top and front views) to form three point density images (PDI). These are then fed into a convolutional neural network (CNN) to classify the targets. The second approach uses an offline target database comprising size and a coarse target silhouette, with classification performed using a few simple rules. Both algorithms do not require knowledge of the aspect angle making them robust when applied in an operational scenario.