Polsar Target Classification Using Polarimetric-Feature-Driven Deep Convolutional Neural Network

Abstract

Deep convolutional neural network (CNN) techniques have been utilized to enhance polarimetric synthetic aperture radar (PolSAR) image classification performance. This work contributes to a current challenge that is how to adapt deep CNN classifier for PolSAR classification with limited training samples while keeping good generalization performance. A polarimetric-feature-driven deep CNN classification scheme is established with both classical roll-invariant polarimetric features and hidden polarimetric features in the rotation domain to drive the proposed deep CNN model. Comparison studies validate the efficiency and superiority of the proposal. For the benchmark AIRSAR data, the proposed method achieves the state-of-the-art classification accuracies. Meanwhile, the convergence speed from the proposed CNN approach is about 2.3 times faster than the normal CNN method. For multi-temporal UAVSAR datasets, the proposed scheme achieves comparably high classification accuracies as the normal CNN method for train-used temporal data, while for train-not-used data it obtains average 4.86% higher overall accuracy than the normal CNN method. Furthermore, the proposed strategy can also produce very promising classification accuracy with very limited training samples.