SAR ATR of Ground Vehicles Based on ESENet

Feng Zhou,Li Wang,Xueru Bai

doi:10.3390/rs11111316

Abstract

In recent studies, synthetic aperture radar (SAR) automatic target recognition (ATR) algorithms that are based on the convolutional neural network (CNN) have achieved high recognition rates in the moving and stationary target acquisition and recognition (MSTAR) dataset. However, in a SAR ATR task, the feature maps with little information automatically learned by CNN will disturb the classifier. We design a new enhanced squeeze and excitation (enhanced-SE) module to solve this problem, and then propose a new SAR ATR network, i.e., the enhanced squeeze and excitation network (ESENet). When compared to the available CNN structures that are designed for SAR ATR, the ESENet can extract more effective features from SAR images and obtain better generalization performance. In the MSTAR dataset containing pure targets, the proposed method achieves a recognition rate of 97.32% and it exceeds the available CNN-based SAR ATR algorithms. Additionally, it has shown robustness to large depression angle variation, configuration variants, and version variants.

Highlights

Synthetic aperture radar (SAR) has played a significant role in surveillance and battlefield reconnaissance, thanks to its all-day, all-weather, and high resolution capability
Cnfounsfiuosniomn mataritxrixofofthEeSEENSEetNuentd. eOr bEvOiCou1.sly, the enhanced squeeze and excitation network (ESENet) outperforms the others and the recognition accuracy is increased by 3% as compared with traditional convolutional neural network (CNN)
Feature extraction plays an important role in the task of synthetic aperture radar (SAR) automatic target recognition (ATR)

Summary

Introduction

Synthetic aperture radar (SAR) has played a significant role in surveillance and battlefield reconnaissance, thanks to its all-day, all-weather, and high resolution capability. It is challenging to identify the targets in SAR images. The MIT Lincoln Laboratory proposed the standard SAR ATR architecture, which consists of three stages: detection, discrimination, and classification [1]. Simple decision rules are used to find the bright pixels in SAR images and indicate the presence of targets. The output of this stage might include targets of interests, and clutters, because the decision stage is far from perfect. On the final classification stage, a classifier is designed to categorize each output image of the discrimination stage as a specific target type

Methods

Results

Conclusion