Anchor-free Convolutional Network with Dense Attention Feature Aggregation for Ship Detection in SAR Images

被引:63
作者
Gao, Fei [1 ]
He, Yishan [1 ]
Wang, Jun [1 ]
Hussain, Amir [2 ]
Zhou, Huiyu [3 ]
机构
[1] Beihang Univ, Sch Elect Informat Engn, Beijing 100191, Peoples R China
[2] Edinburgh Napier Univ, Sch Comp, Cognit Big Data & Cyber Informat CogBID Lab, Edinburgh EH10 5DT, Midlothian, Scotland
[3] Univ Leicester, Dept Informat, Leicester LE1 7RH, Leics, England
基金
英国工程与自然科学研究理事会; 中国国家自然科学基金;
关键词
ship detection; convolutional neural networks (CNN); synthetic aperture radar (SAR); anchor-free; feature aggregation; attention mechanism; TARGET DETECTION; ALGORITHM;
D O I
10.3390/rs12162619
中图分类号
X [环境科学、安全科学];
学科分类号
08 ; 0830 ;
摘要
In recent years, with the improvement of synthetic aperture radar (SAR) imaging resolution, it is urgent to develop methods with higher accuracy and faster speed for ship detection in high-resolution SAR images. Among all kinds of methods, deep-learning-based algorithms bring promising performance due to end-to-end detection and automated feature extraction. However, several challenges still exist: (1) standard deep learning detectors based on anchors have certain unsolved problems, such as tuning of anchor-related parameters, scale-variation and high computational costs. (2) SAR data is huge but the labeled data is relatively small, which may lead to overfitting in training. (3) To improve detection speed, deep learning detectors generally detect targets based on low-resolution features, which may cause missed detections for small targets. In order to address the above problems, an anchor-free convolutional network with dense attention feature aggregation is proposed in this paper. Firstly, we use a lightweight feature extractor to extract multiscale ship features. The inverted residual blocks with depth-wise separable convolution reduce the network parameters and improve the detection speed. Secondly, a novel feature aggregation scheme called dense attention feature aggregation (DAFA) is proposed to obtain a high-resolution feature map with multiscale information. By combining the multiscale features through dense connections and iterative fusions, DAFA improves the generalization performance of the network. In addition, an attention block, namely spatial and channel squeeze and excitation (SCSE) block is embedded in the upsampling process of DAFA to enhance the salient features of the target and suppress the background clutters. Third, an anchor-free detector, which is a center-point-based ship predictor (CSP), is adopted in this paper. CSP regresses the ship centers and ship sizes simultaneously on the high-resolution feature map to implement anchor-free and nonmaximum suppression (NMS)-free ship detection. The experiments on the AirSARShip-1.0 dataset demonstrate the effectiveness of our method. The results show that the proposed method outperforms several mainstream detection algorithms in both accuracy and speed.
引用
收藏
页数:25
相关论文
共 57 条
[1]   A New CFAR Ship Detection Algorithm Based on 2-D Joint Log-Normal Distribution in SAR Images [J].
Ai, Jiaqiu ;
Qi, Xiangyang ;
Yu, Weidong ;
Deng, Yunkai ;
Liu, Fan ;
Shi, Li .
IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2010, 7 (04) :806-810
[2]   Detection and Discrimination of Ship Targets in Complex Background From Spaceborne ALOS-2 SAR Images [J].
Ao, Wei ;
Xu, Feng ;
Li, Yongchen ;
Wang, Haipeng .
IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2018, 11 (02) :536-550
[3]   Ship Detection Based on YOLOv2 for SAR Imagery [J].
Chang, Yang-Lang ;
Anagaw, Amare ;
Chang, Lena ;
Wang, Yi Chun ;
Hsiao, Chih-Yu ;
Lee, Wei-Hong .
REMOTE SENSING, 2019, 11 (07)
[4]   A Deep Neural Network Based on an Attention Mechanism for SAR Ship Detection in Multiscale and Complex Scenarios [J].
Chen, Chen ;
He, Chuan ;
Hu, Changhua ;
Pei, Hong ;
Jiao, Licheng .
IEEE ACCESS, 2019, 7 :104848-104863
[5]  
Chen Kai, 2019, arXiv:1906.07155
[6]  
Crisp D.J., 2004, The State of the Art in Ship Detection in Synthetic Aperture Radar Imagery
[7]   Dense Attention Pyramid Networks for Multi-Scale Ship Detection in SAR Images [J].
Cui, Zongyong ;
Li, Qi ;
Cao, Zongjie ;
Liu, Nengyuan .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2019, 57 (11) :8983-8997
[8]  
Deng J, 2009, PROC CVPR IEEE, P248, DOI 10.1109/CVPRW.2009.5206848
[9]   Learning Deep Ship Detector in SAR Images From Scratch [J].
Deng, Zhipeng ;
Sun, Hao ;
Zhou, Shilin ;
Zhao, Juanping .
IEEE TRANSACTIONS ON GEOSCIENCE AND REMOTE SENSING, 2019, 57 (06) :4021-4039
[10]   Ship Detection Using a Fully Convolutional Network with Compact Polarimetric SAR Images [J].
Fan, Qiancong ;
Chen, Feng ;
Cheng, Ming ;
Lou, Shenlong ;
Xiao, Rulin ;
Zhang, Biao ;
Wang, Cheng ;
Li, Jonathan .
REMOTE SENSING, 2019, 11 (18)