一种基于稀疏化核方法的红外强杂波背景抑制算法

朱斌, 樊祥, 程正东, 王迪, 方义强, 陈晓斯

电子学报 ›› 2015, Vol. 43 ›› Issue (4) : 716-721.

PDF(1861 KB)
PDF(1861 KB)
电子学报 ›› 2015, Vol. 43 ›› Issue (4) : 716-721. DOI: 10.3969/j.issn.0372-2112.2015.04.013
学术论文

一种基于稀疏化核方法的红外强杂波背景抑制算法

  • 朱斌1,2, 樊祥1,2, 程正东1,2, 王迪1, 方义强1, 陈晓斯1
作者信息 +

An IR Strong Clutter Background Suppression Algorithm Based on Sparse Kernel Method

  • ZHU Bin1,2, FAN Xiang1,2, CHENG Zheng-dong1,2, WANG Di1, FANG Yi-qiang1, CHEN Xiao-si1
Author information +
文章历史 +

摘要

杂波背景抑制一直是红外弱小目标检测面临的难题.背景抑制可分为背景预测和差分滤波两步.针对强杂波背景呈现非线性分布的特征,提出了一种基于稀疏化核递推最小二乘(KRLS)算法的非线性背景抑制算法.算法采用监督学习模型,使用序列图像作为训练样本.通过稀疏化控制学习函数的复杂度并剔除冗余信息,不但可以提高学习机器的推广能力,还可以降低运算量.使用真实红外图像对算法进行了测试,并分析了算法参数.实验结果表明:算法可自适应预测不同类型的强杂波背景,并有效抑制背景杂波.

Abstract

Clutter background suppression is always a difficulty of infrared (IR) dim and point target detection.Background suppression is divided into background estimation and difference filtering.Aiming at the nonlinear distribution of strong clutter background,a spares kernel recursive least squares (KRLS) based nonlinear background suppression algorithm is proposed.This method uses sequence images as training sample in supervised learning model.The complexity of learned function is controlled,and the redundant information is discarded by sparsification.In this way,the generalization of learning machine can be enhanced;moreover,the computational burden can be reduced.In the experiments,real IR images are used to test the algorithm,and the parameters are analyzed.Experimental results show that different kinds of strong clutter background can be estimated,and then be suppressed.

关键词

红外背景抑制 / 强杂波 / 背景预测 / 稀疏 / 核递推最小二乘

Key words

IR background suppression / strong clutter / background estimation / sparse / kernel recursive least squares

引用本文

导出引用
朱斌, 樊祥, 程正东, 王迪, 方义强, 陈晓斯. 一种基于稀疏化核方法的红外强杂波背景抑制算法[J]. 电子学报, 2015, 43(4): 716-721. https://doi.org/10.3969/j.issn.0372-2112.2015.04.013
ZHU Bin, FAN Xiang, CHENG Zheng-dong, WANG Di, FANG Yi-qiang, CHEN Xiao-si. An IR Strong Clutter Background Suppression Algorithm Based on Sparse Kernel Method[J]. Acta Electronica Sinica, 2015, 43(4): 716-721. https://doi.org/10.3969/j.issn.0372-2112.2015.04.013
中图分类号: TN911.73   

参考文献

[1] Stephen J Searle.Background modeling and target segmentation via modified Kalman filtering[A].Proc SPIE 5428,Signal and Data Processing of Small Target 2004[C].SPIE,2004,13-24.
[2] 朱斌,樊祥,马东辉,等.核最小二乘算法检测红外点目标[J].光电工程,2009,36(9):29-34. Bin Zhu,Xiang Fan,Donghui Ma,et al.Infrared point target detection based on kernel least squares algorithm[J].Opto-Electronic Engineering,2009,36(9):29-34.(in Chinese)
[3] Bin Zhu,Xiang Fan,Donghui Ma,et al.Sequence IR images background estimation algorithm based on kernel exponential weighted least squares[A].Proc SPIE 7495,MIPPR 2009:Automatic Target Recognition and Image Analysis[C].SPIE,2009,doi:10.1117/12.832993.
[4] Soni T,Zeidler J R,Ku H.Performance evaluation of 2-D adaptive prediction filters for detection of small objects in image data[J].IEEE Trans on Image Processing,1993,2(3):327-340.
[5] V N Vapnik.Statistical Learning Theory[M].New York:John Wiley & Sons,Inc.1998.
[6] Yaakov Engel,Shie Mannor,Ron Meir.The kernel recursive least-squares algorithm[J].IEEE Trans on Signal Processing,2004,52(8):2275-2285.
[7] Bin Zhu,Zhengdong Cheng,Xiang Fan,et al.Online sparse IR background estimation via KRLS[A].2010 IEEE International Conference on Information and Automation(ICIA)[C].IEEE,2010.1123-1127.
[8] John Shawe-Taylor,Nello Cristianini.Kernel Methods for Pattern Analysis[M].Cambridge University Press,2004.
[9] Marcelo R P Ferreira,Francisco de A T de Carvalho.Kernel-based hard clustering methods in the feature space with automatic variable weighting[J].Pattern Recognition,2014,47:3082-3095.

基金

国家自然科学基金 (No.61307025); 安徽省自然科学基金 (No.1308085QF122)

PDF(1861 KB)

1670

Accesses

0

Citation

Detail

段落导航
相关文章

/