A novel target detection method for SAR images based on shadow proposal and saliency analysis

Fei Gao, Jialing You, Jun Wang, Jinping Sun, Erfu Yang, Huiyu Zhou

Research output: Contribution to journalArticle

  • 2 Citations

Abstract

Conventional synthetic aperture radar (SAR) based target detection methods generally use high intensity pixels in the pre-screening stage while ignoring shadow information. Furthermore, they cannot accurately extract the target area and also have poor performance in cluttered environments. To solve this problem, a novel SAR target detection method which combines shadow proposal and saliency analysis is presented in this paper. The detection process is divided into shadow proposal, saliency detection and One-Class Support Vector Machine (OC-SVM) screening stages. In the shadow proposal stage, localizing targets is performed rst with the detected shadow regions to generate proposal chips that may contain potential targets. Then saliency detection is conducted to extract salient regions of the proposal chips using local spatial autocorrelation and signicance tests. Afterwards, in the last stage, the OC-SVM is employed to identify the real targets from the salient regions. Experimental results show that the proposed saliency detection method possesses higher detection accuracy than several state of the art methods on SAR images. Furthermore, the proposed SAR target detection method is demonstrated to be robust under dierent imaging environments. to extract salient regions of the proposal chips using local spatial autocorrelation and signicance tests. Afterwards, in the last stage, the OC-SVM is employed
LanguageEnglish
Pages220-231
Number of pages12
JournalNeurocomputing
Volume267
Early online date8 Jun 2017
DOIs
StatePublished - 6 Dec 2017

Fingerprint

Radar
Synthetic aperture radar
Target tracking
Support vector machines
Autocorrelation
Spatial Analysis
Screening
Pixels
Imaging techniques
Support Vector Machine

Keywords

  • synthetic aperture radar (SAR) target detection
  • shadow
  • saliency detection
  • local spatial autocorrelation,
  • visual features
  • one-class SVM (OC-SVM)

Cite this

Gao, Fei ; You, Jialing ; Wang, Jun ; Sun, Jinping ; Yang, Erfu ; Zhou, Huiyu. / A novel target detection method for SAR images based on shadow proposal and saliency analysis. In: Neurocomputing. 2017 ; Vol. 267. pp. 220-231
@article{6fb7c7afa05a47d89767206c1be5d522,
title = "A novel target detection method for SAR images based on shadow proposal and saliency analysis",
abstract = "Conventional synthetic aperture radar (SAR) based target detection methods generally use high intensity pixels in the pre-screening stage while ignoring shadow information. Furthermore, they cannot accurately extract the target area and also have poor performance in cluttered environments. To solve this problem, a novel SAR target detection method which combines shadow proposal and saliency analysis is presented in this paper. The detection process is divided into shadow proposal, saliency detection and One-Class Support Vector Machine (OC-SVM) screening stages. In the shadow proposal stage, localizing targets is performed rst with the detected shadow regions to generate proposal chips that may contain potential targets. Then saliency detection is conducted to extract salient regions of the proposal chips using local spatial autocorrelation and signicance tests. Afterwards, in the last stage, the OC-SVM is employed to identify the real targets from the salient regions. Experimental results show that the proposed saliency detection method possesses higher detection accuracy than several state of the art methods on SAR images. Furthermore, the proposed SAR target detection method is demonstrated to be robust under dierent imaging environments. to extract salient regions of the proposal chips using local spatial autocorrelation and signicance tests. Afterwards, in the last stage, the OC-SVM is employed",
keywords = "synthetic aperture radar (SAR) target detection, shadow, saliency detection, local spatial autocorrelation,, visual features, one-class SVM (OC-SVM)",
author = "Fei Gao and Jialing You and Jun Wang and Jinping Sun and Erfu Yang and Huiyu Zhou",
year = "2017",
month = "12",
day = "6",
doi = "10.1016/j.neucom.2017.06.004",
language = "English",
volume = "267",
pages = "220--231",
journal = "Neurocomputing",
issn = "0925-2312",

}

A novel target detection method for SAR images based on shadow proposal and saliency analysis. / Gao, Fei; You, Jialing; Wang, Jun; Sun, Jinping; Yang, Erfu; Zhou, Huiyu.

In: Neurocomputing, Vol. 267, 06.12.2017, p. 220-231.

Research output: Contribution to journalArticle

TY - JOUR

T1 - A novel target detection method for SAR images based on shadow proposal and saliency analysis

AU - Gao,Fei

AU - You,Jialing

AU - Wang,Jun

AU - Sun,Jinping

AU - Yang,Erfu

AU - Zhou,Huiyu

PY - 2017/12/6

Y1 - 2017/12/6

N2 - Conventional synthetic aperture radar (SAR) based target detection methods generally use high intensity pixels in the pre-screening stage while ignoring shadow information. Furthermore, they cannot accurately extract the target area and also have poor performance in cluttered environments. To solve this problem, a novel SAR target detection method which combines shadow proposal and saliency analysis is presented in this paper. The detection process is divided into shadow proposal, saliency detection and One-Class Support Vector Machine (OC-SVM) screening stages. In the shadow proposal stage, localizing targets is performed rst with the detected shadow regions to generate proposal chips that may contain potential targets. Then saliency detection is conducted to extract salient regions of the proposal chips using local spatial autocorrelation and signicance tests. Afterwards, in the last stage, the OC-SVM is employed to identify the real targets from the salient regions. Experimental results show that the proposed saliency detection method possesses higher detection accuracy than several state of the art methods on SAR images. Furthermore, the proposed SAR target detection method is demonstrated to be robust under dierent imaging environments. to extract salient regions of the proposal chips using local spatial autocorrelation and signicance tests. Afterwards, in the last stage, the OC-SVM is employed

AB - Conventional synthetic aperture radar (SAR) based target detection methods generally use high intensity pixels in the pre-screening stage while ignoring shadow information. Furthermore, they cannot accurately extract the target area and also have poor performance in cluttered environments. To solve this problem, a novel SAR target detection method which combines shadow proposal and saliency analysis is presented in this paper. The detection process is divided into shadow proposal, saliency detection and One-Class Support Vector Machine (OC-SVM) screening stages. In the shadow proposal stage, localizing targets is performed rst with the detected shadow regions to generate proposal chips that may contain potential targets. Then saliency detection is conducted to extract salient regions of the proposal chips using local spatial autocorrelation and signicance tests. Afterwards, in the last stage, the OC-SVM is employed to identify the real targets from the salient regions. Experimental results show that the proposed saliency detection method possesses higher detection accuracy than several state of the art methods on SAR images. Furthermore, the proposed SAR target detection method is demonstrated to be robust under dierent imaging environments. to extract salient regions of the proposal chips using local spatial autocorrelation and signicance tests. Afterwards, in the last stage, the OC-SVM is employed

KW - synthetic aperture radar (SAR) target detection

KW - shadow

KW - saliency detection

KW - local spatial autocorrelation,

KW - visual features

KW - one-class SVM (OC-SVM)

UR - http://www.sciencedirect.com/science/journal/09252312

U2 - 10.1016/j.neucom.2017.06.004

DO - 10.1016/j.neucom.2017.06.004

M3 - Article

VL - 267

SP - 220

EP - 231

JO - Neurocomputing

T2 - Neurocomputing

JF - Neurocomputing

SN - 0925-2312

ER -