电子学报 ›› 2016, Vol. 44 ›› Issue (11): 2633-2638.DOI: 10.3969/j.issn.0372-2112.2016.11.010

• 学术论文 • 上一篇    下一篇

联合表示求解二元假设模型的高光谱目标检测

凌强, 黄树彩, 韦道知, 吴潇   

  1. 空军工程大学防空反导学院, 陕西西安 710051
  • 收稿日期:2015-06-04 修回日期:2015-08-24 出版日期:2016-11-25
    • 作者简介:
    • 凌强,男,1990年生于湖南双峰.硕士研究生.研究方向为高光谱目标检测、弹道目标检测.E-mail:lq910131@163.com;黄树彩,男,1967年生于湖北黄梅,博士,教授.研究方向为模式识别、目标检测与跟踪.
    • 基金资助:
    • 航空科学基金 (No.20130196004)

Collaborative Representation-Based Binary Hypothesis Model for Hyperspectral Target Detection

LING Qiang, HUANG Shu-cai, WEI Dao-zhi, WU Xiao   

  1. Air and Missile Defense College, Air Force Engineering University, Xi'an, Shaanxi 710051, China
  • Received:2015-06-04 Revised:2015-08-24 Online:2016-11-25 Published:2016-11-25
    • Supported by:
    • Aeronautical Science Foundation of China, ASFC (No.20130196004)

摘要:

针对稀疏表示目标检测理论中稀疏度难以确定的问题,本文将联合表示应用于目标检测,提出了一种新颖的目标检测算法,并给出了该算法的非线性形式.其核心思想是:背景像元的光谱能够被其周围背景像元的光谱(背景字典)线性表示,而目标像元的光谱只能被其周围背景像元的光谱和目标先验光谱(联合字典)线性表示.该算法首先用背景字典和联合字典分别对待检测像元进行联合表示,然后比较两次联合表示的重构误差确定像元类别.通过真实的高光谱图像进行验证,结果表明,与其它目标检测算法相比,该算法具有较好的检测性能.

关键词: 目标检测, 联合表示, 核联合表示, 高光谱图像

Abstract:

In order to solve the problem of setting sparsity level in sparse representation-based target detection algorithms,this paper proposes a novel collaborative representation-based algorithm for hyperspectral target detection,and then extends it into a kernel version.The key idea is that a background pixel can be approximately represented as a linear combination of its surrounding neighbors (background dictionary),while a target pixel can only be approximately represented as a linear combination of its surrounding neighbors and the prior target spectrums (union dictionary).First the unknown pixel is collaboratively represented by the background dictionary and union dictionary,respectively.Then targets can be determined by comparing the reconstruction residuals.Experimental results on real hyperspectral data set demonstrate the effectiveness of our proposed detector as well as its kernel version when compared with other algorithms.

Key words: target detection, collaborative representation, kernel collaborative representation, hyperspectral imagery

中图分类号: