Deep neural network design for SAR/ISAR-based automatic target recognition

The automatic recognition of targets with radar is an ongoing research field for many years already. Since 2014, a new methodology based on deep neural networks is becoming more established within this field. This chapter gives a short overview with some examples of this short history of target recognition using deep learning (DL) and some comparative results with a basic implementation of a convolutional neural network (CNN). This network is applied to the commonly used Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset and to an inverse synthetic aperture radar (ISAR) dataset measured with the Tracking and Imaging Radar (TIRA) of Fraunhofer FHR. Chapter Contents: ⢠8.1 Introduction ⢠8.2 Deep learning methods used for target recognition ⢠8.3 Datasets ⢠8.3.1 MSTAR SAR dataset ⢠8.3.2 TIRA ISAR dataset ⢠8.3.3 Artificial training data ⢠8.4 Classification system ⢠8.4.1 CNN structure ⢠8.4.2 Training algorithm ⢠8.4.2.1 Stocha stic diagonal Levenberg-Marquardt algorithm ⢠8.4.2.2 Adam algorithm ⢠8.4.3 Regularization methods ⢠8.5 Experimental results ⢠8.5.1 CNN structure ⢠8.5.2 Training method and cost function ⢠8.5.3 Regularization methods ⢠8.5.4 Artificial training data ⢠8.5.5 Use of SVMs as classifier ⢠8.6 Summary and conclusion ⢠References.