When Deep Learning Meets Multi-Task Learning in SAR ATR: Simultaneous Target Recognition and Segmentation

Chenwei Wang,Yulin Huang,Haiguang Yang,Junjie Wu,Zhiyong Wang,Jifang Pei,Jianyu Yang

doi:10.3390/rs12233863

Abstract

With the recent advances of deep learning, automatic target recognition (ATR) of synthetic aperture radar (SAR) has achieved superior performance. By not being limited to the target category, the SAR ATR system could benefit from the simultaneous extraction of multifarious target attributes. In this paper, we propose a new multi-task learning approach for SAR ATR, which could obtain the accurate category and precise shape of the targets simultaneously. By introducing deep learning theory into multi-task learning, we first propose a novel multi-task deep learning framework with two main structures: encoder and decoder. The encoder is constructed to extract sufficient image features in different scales for the decoder, while the decoder is a tasks-specific structure which employs these extracted features adaptively and optimally to meet the different feature demands of the recognition and segmentation. Therefore, the proposed framework has the ability to achieve superior recognition and segmentation performance. Based on the Moving and Stationary Target Acquisition and Recognition (MSTAR) dataset, experimental results show the superiority of the proposed framework in terms of recognition and segmentation.

Highlights

Synthetic aperture radar (SAR) is an important microwave remote sensing system in the domains of military and civilian applications
The stability and effectiveness of the proposed network architecture are evaluated in the variances of the depression angle, target configuration and version, which are denoted as extended operating conditions (EOCs)-D, EOC-C and EOC-V, respectively
The synthetic aperture radar (SAR) images are extremely sensitive to the variance of the depression angle, so it is important to evaluate the performance of the proposed network architecture at the variance of depression angle, EOC-D

Summary

Introduction

Synthetic aperture radar (SAR) is an important microwave remote sensing system in the domains of military and civilian applications. With the high-resolution coherent imaging capability of all weather and all day penetration, it can obtain more distinct information than optical sensors, infrared sensors, etc. It is able to acquire abundant backscattering characteristics of the targets. These backscattering characteristics contain unique identifying information of target attributes, which is often difficult to accurately interpret from the perspective of human vision. It is usually a hard task to accomplish real-time processing when the size and number of SAR images are increasing. SAR automatic target recognition (ATR) has become one of the most crucial and challenging issues in SAR application

Methods

Results

Conclusion