改进时空滤波的红外弱小目标检测

樊香所; 范锦龙; 文良华; 徐智勇

改进时空滤波的红外弱小目标检测

樊香所^{1, 2,},
范锦龙^3, ,,
文良华¹,
徐智勇⁴

1.
宜宾学院智能制造学部, 四川宜宾 644600
2.
广西科技大学, 广西土方机械协同创新中心, 广西柳州 545006
3.
国家卫星气象中心, 北京 100081
4.
中国科学院光电技术研究所, 四川成都 610209

基金项目:

国家自然科学基金项目 62001129

广西自然科学基金 2021GXNSFBA075029

国家自然科学基金项目 61975171

详细信息

作者简介:
樊香所（1987-），男，博士，研究方向为计算机视觉与图像处理，E-mail: 100002085@gxust.edu.cn

通讯作者:
范锦龙（1975-），男，副研究员，主要从事遥感图像处理，E-mail: fanjl@cma.gov.cn

中图分类号: TP751
计量
- 文章访问数: 208
- HTML全文浏览量: 64
- PDF下载量: 49
出版历程
- 收稿日期: 2021-01-10
- 修回日期: 2021-06-27
- 刊出日期: 2022-05-19

Infrared Dim-Small Target Detection Based on Improved Spatio-Temporal Filtering

1.
Division of Intelligent Manufacturing, Yibin University, Yibin 644600, China
2.
Guangxi Earthmoving Machinery Collaborative Innovation Center, Guangxi University of Science and Technology, Liuzhou 545006, China
3.
National Satellite Meteorological Center, Beijing 100081, China
4.
Institute of Optics and Electronics, Chinese Academy of Sciences, Chengdu 610209, China

摘要

摘要: 为了有效解决动态背景变化导致弱小目标检测率低的问题，文中提出了改进时空滤波的红外弱小目标检测算法。首先在分析红外图像成像特性的基础上，针对目标区、背景区和边缘轮廓区不同梯度特性的差异，提出改进的各向异性空域滤波算法，该算法充分利用空间域的梯度信息来构建不同方向的扩散滤波函数，并结合图像不同特性的梯度差异选取扩散函数值最小的两个方向的均值作为时域滤波结果，以最大限度地保留目标信号；接着为有效增强弱小目标的能量，针对高阶累积量仅利用像元点时域信息来构建能量增强的不足，提出了一种结合时空邻域块的能量增强算法，实验表明，本文提出的算法能有效提升动态场景下的弱小目标的检测能力。
- 弱小目标 /
- 改进各向异性 /
- 能量增强 /
- 目标检测
Abstract: To effectively solve the problem of low detection rates of dim and small targets caused by dynamic background changes, a detection method based on spatio-temporal filtering is proposed in this paper. Based on an analysis of the imaging characteristics of infrared images, an improved anisotropic spatial filtering algorithm is proposed to evaluate the difference in various gradient characteristics of the target area, background area, and edge contour area. The algorithm fully utilizes the gradient information in the spatial domain to construct the diffusion filter function in different directions. According to the gradient difference in various characteristics of the image, the mean value of the two directions with the smallest value of the diffusion function is selected as the result of spatial filtering to retain the target signal to the maximum extent. To effectively enhance the energy of dim and small targets and address the shortcomings of high-order cumulants that only use the temporal domain information of pixel points for energy enhancement, an energy enhancement algorithm based on spatial-temporal neighborhood blocks is proposed. Experimental results reveal that the proposed algorithm can effectively enhance the detection of dim and small targets in dynamically changing scenes.
- dim and small target /
- improve anisotropic filtering /
- energy enhancement /
- targets detection

HTML全文

图 1 图像中不同组分的梯度差异

Figure 1. The gradient difference of different components in the image

下载: 全尺寸图片幻灯片

图 2 步长k与LSNR间的关系

Figure 2. The relationship between k and LSNR

下载: 全尺寸图片幻灯片

图 3 场景1不同算法的检测结果

Figure 3. Detection results of different algorithms in scene 1

下载: 全尺寸图片幻灯片

图 4 场景2不同算法的检测结果

Figure 4. Detection results of different algorithms in scene 2

下载: 全尺寸图片幻灯片

图 5 场景3不同算法的检测结果

Figure 5. Detection results of different algorithms in scene 3

下载: 全尺寸图片幻灯片

图 6 3个场景ROC曲线

Figure 6. ROC curves of different algorithms in three scenes

下载: 全尺寸图片幻灯片

表 1 序列图像信息

Table 1 Information of sequence images

Image sequence	Frame length/frame	Image size	Target size
1	114	278×246	3×3
2	50	180×180	3×3
3	54	180×180	3×3

下载: 导出CSV

表 2 不同算法获取的背景建模结果

Table 2 Results obtained by different algorithms

Scene	Evaluating indicator	ITH	LRR	IPI	AF	Proposed method
1	AGV	16.44	32.44	44.88	14.57	58.89
1	LSNR/dB	2.61	2.26	2.77	2.10	3.38
2	AGV	26.89	27.67	30.22	22.78	48.56
2	LSNR/dB	2.68	2.62	2.59	2.39	3.27
3	AGV	17.33	33.56	34.89	17.18	72.44
3	LSNR/dB	2.79	3.11	3.23	2.39	3.20
4	AGV	21.33	22.78	28.33	17.63	73.67
4	LSNR/dB	2.27	2.87	2.92	2.70	3.24
5	AGV	24.89	25.11	27.44	20.71	97.00
5	LSNR/dB	2.23	2.49	2.62	2.56	3.14

下载: 导出CSV

表 3 两种算法的增强结果

Table 3 Enhancement comparison of two algorithms

Difference image		HOC		Proposed method
AGV	LSNR/dB	AGV	LSNR/dB	AGV	LSNR/dB
72	3.20	102	5.75	225	12.67

下载: 导出CSV

参考文献(13)

[1]	樊香所. 序列图像弱小目标检测与跟踪算法研究[D]. 成都: 电子科技大学, 2019. FAN Xiangsuo. Dim and Small Targets Detection and Tracking Algorithms in Sequence Image[D]. Chengdu: University of Electronic Science and Technology of China, 2019.
[2]	Bae T W, Kim Y C, Ahn S H, et al. An efficient two dimensional least mean square based on block statistics for small target detection[J]. Journal of Infrared, Millimeter, and Terahertz Waves, 2009, 30(10): 1092-1101. DOI: 10.1007/s10762-009-9530-6
[3]	BAI X Z, ZHOU F G, JIN T. Enhancement of dim small target through modified top-hat transformation under the condition of heavy clutter[J]. Signal Processing, 2010, 90(1): 1643-1654.
[4]	秦翰林, 周慧鑫, 刘上乾, 等. 基于双边滤波的弱小目标背景抑制[J]. 强激光与粒子束, 2009, 21(1): 25-28. https://www.cnki.com.cn/Article/CJFDTOTAL-QJGY200901007.htm QIN H, ZHOU H, LIU S, et al. Dim and small target background suppression using bilateral filtering[J]. High Power Laser and Particle Beams, 2009, 21(1): 25-28. https://www.cnki.com.cn/Article/CJFDTOTAL-QJGY200901007.htm
[5]	严高师, 毕务忠. 基于区域奇异性滤波的小目标检测[J]. 光学技术, 2006, 33(2): 163-165, 169. DOI: 10.3321/j.issn:1002-1582.2006.02.003 YAN G S, BI W Z. Detection algorithm of small target based on regional singularity filter[J]. Optical Technology, 2006, 33(2): 163-165, 169. DOI: 10.3321/j.issn:1002-1582.2006.02.003
[6]	连可, 王厚军, 李丹. 基于红外目标局部灰度特性分析的管道滤波方法[J]. 弹箭与制导学报, 2011, 31(4): 200-206. DOI: 10.3969/j.issn.1673-9728.2011.04.057 LIAN K, WANG H J, LI D. Pipeline filtering method based on feature analysis of local grey level of small infrared target[J]. Journal of Projectiles, Rockets, Missiles and Guidance, 2011, 31(4): 200-203, 206. DOI: 10.3969/j.issn.1673-9728.2011.04.057
[7]	Oliver N M, Rosario B, Pentland A P. A Bayesian computer vision system for modeling human interactions[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2000, 22(8): 831-843. DOI: 10.1109/34.868684
[8]	Bouwmans T, Baf F E, Vachon B. Background modeling using mixture of Gaussians for foreground detection - a survey [J]. Recent Patents on Computer Science, 2008, 1(3): 219-237. DOI: 10.2174/2213275910801030219
[9]	何玉杰, 李敏, 张金利, 等. 基于低秩三分解的红外图像杂波抑制[J]. 光学与精密工程, 2015, 23(7): 2069-2078. https://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201507034.htm HE Y J, LI M, ZHANG J L, et al. Clutter suppression of infrared image based on three-component low rank matrix decomposition[J]. Optics and Precision Engineering, 2015, 23(7): 2069-2078 https://www.cnki.com.cn/Article/CJFDTOTAL-GXJM201507034.htm
[10]	GUO J, WU Y, DAI Y. Small target detection based on reweighted infrared patch-image model[J]. IET Image Processing, 2018, 12(1): 70-79.
[11]	陆福星, 李夜金, 陈忻, 等. 基于Top-Hat变换的PM模型弱小目标检测[J]. 系统工程与电子技术, 2018, 40(7): 1417-1422. https://www.cnki.com.cn/Article/CJFDTOTAL-XTYD201807001.htm LU F X, LI Y J, CHEN X, et al. Weak target detection for PM model based on Top-hat transform[J]. Systems Engineering and Electronics, 2018, 40(7): 1417-1422. https://www.cnki.com.cn/Article/CJFDTOTAL-XTYD201807001.htm
[12]	周慧鑫, 赵营, 秦翰林, 等. 多尺度各向异性扩散方程的红外弱小目标检测算法[J]. 光子学报, 2015, 44(9): 146-150. https://www.cnki.com.cn/Article/CJFDTOTAL-GZXB201509027.htm ZHOU H X, ZHAO Y, QIN H L, et al. Infrared dim and small target detection algorithm based on multi-scale anisotropic diffusion equation[J]. Acta Photonica Sinica, 2015, 44(9): 146-150. https://www.cnki.com.cn/Article/CJFDTOTAL-GZXB201509027.htm
[13]	唐意东, 黄树彩, 钟宇, 等. 基于形态学和高阶统计量的弱小运动目标检测[J]. 现代防御技术, 2016, 44(2): 151-156. DOI: 10.3969/j.issn.1009-086x.2016.02.024 TANG Y D, HUANG S C, ZHONG Y, et al. Moving dim target detection based on morphology and high-order statistics in infrared image[J]. Modern Defense Technology, 2016, 44(2): 151-156. DOI: 10.3969/j.issn.1009-086x.2016.02.024