系统工程与电子技术 ›› 2021, Vol. 43 ›› Issue (10): 2766-2774.doi: 10.12305/j.issn.1001-506X.2021.10.08

• 电磁散射与逆散射研究新进展专栏 • 上一篇    下一篇

基于Wishart距离和超像素的极化SAR图像车辆检测

代晓康1, 殷君君1,*, 杨健2   

  1. 1. 北京科技大学计算机与通信工程学院, 北京 100083
    2. 清华大学电子工程系, 北京 100084
  • 收稿日期:2021-02-22 出版日期:2021-10-01 发布日期:2021-11-04
  • 通讯作者: 殷君君
  • 作者简介:代晓康(1996—), 男, 硕士研究生, 主要研究方向为极化SAR图像目标检测|殷君君(1983—), 女, 副教授,博士, 主要研究方向为雷达极化应用的基础理论、极化合成孔径雷达图像理解、图像分割与地物分类、目标检测等|杨健(1965—), 男, 教授, 博士研究生导师, 博士,主要研究方向为雷达极化的遥感应用理论、最优极化、数学建模与模糊集理论
  • 基金资助:
    国家自然科学基金(61771043);中央高校基本科研业务费专项资金资助课题(FRF-GF-20-17B);中央高校基本科研业务费专项资金资助课题(FRF-GF-19-017B)

Vehicle detection based on Wishart distance and superpixel in polarimetric SAR image

Xiaokang DAI1, Junjun YIN1,*, Jian YANG2   

  1. 1. School of Computer and Communication Engineering, University of Science and Technology Beijing, Beijing 100083, China
    2. Department of Electronic Engineering, Tsinghua University, Beijing 100084, China
  • Received:2021-02-22 Online:2021-10-01 Published:2021-11-04
  • Contact: Junjun YIN

摘要:

针对极化合成孔径雷达(synthetic aperture radar, SAR)在城市区域复杂地物条件下的密集车辆目标检测问题, 提出了一种结合超像素分割和Wishart分类器的非监督目标检测方法。首先,根据不同地物的极化散射特征检测出建筑物。然后,利用不包含建筑物的Wishart分类器和超像素分割获得目标的形态信息。接着,利用包含建筑物的Wishart分类器获得目标中心点。最后,通过区域生长对二者进行信息融合并完成目标检测任务。基于X波段的机载极化SAR数据表明, 所提算法不仅可以对密集目标进行区分和定位, 并且目标形态保持完整; 相比于传统方法, 目标检测与虚警鉴别性能得到较大提升。

关键词: 车辆检测, 超像素分割, 极化合成孔径雷达, Wishart分类

Abstract:

Aiming at the problem of dense vehicle target detection in polarimetric synthetic aperture radar (SAR) from urban areas under complex scenarios, an unsupervised target detection method that combines the superpixel segmentation and the Wishart classifier is proposed. Firstly, the buildings are detected based on the different polarimetric scattering characteristics of ground objects. Then, the morphological information of the target is obtained by the Wishart classifier without buildings and the superpixel segmentation. After that, the center points of the target are obtained by the Wishart classifier with buildings. Finally, the region growing procedure is used to fuse the information obtained by above-mentio ned classifiers to complete the target detection task. Experimental results implemented on X-band airborne polarimetric SAR data illustrate that the proposed method can not only distinguish and locate dense targets, but also keep the shape of targets. Compared with traditional methods, the performance of target detection and false alarm discrimiantion is improved.

Key words: vehicle detection, superpixel segmentation, polarimetric synthetic aperture radar(SAR), Wishart classifier

中图分类号: