基于局部熵参考预处理的RPCA红外小目标检测

薛锡瑞; 黄树彩; 马佳顺; 李宁

基于局部熵参考预处理的RPCA红外小目标检测

空军工程大学，陕西西安 710051

基金项目:

国家自然科学基金 61573374

详细信息

作者简介:
薛锡瑞（1997-），男，硕士研究生，研究方向为红外弱小目标检测，E-mail：13186032316@163.com

中图分类号: TP751.1
计量
- 文章访问数: 294
- HTML全文浏览量: 92
- PDF下载量: 54
- 被引次数: 0
出版历程
- 收稿日期: 2020-10-06
- 修回日期: 2020-11-06
- 刊出日期: 2021-07-01

RPCA Infrared Small Target Detection Based on Local Entropy Reference in Preprocessing

Air Force Engineering University, Xi'an 710051, China

摘要

摘要: 以图像非局部相似性为基础，利用图像分块重组以获得低秩块图像，是将鲁棒主成分分析算法（robust principal component analysis，RPCA）应用到单帧图像红外小目标检测的基本方法。本文介绍了RPCA算法在单帧图像红外小目标检测的应用流程，分析了不同图像背景下各种分块方法的影响。为解决复杂背景下图像分块窗口和滑动步长难以选择的问题，提出了以图像分块最小局部熵的较大值为参考的选择方法。实验结果表明，通过计算图像的分块局部熵，以最小局部熵的较大值为参考，选择RPCA算法预处理方案，能使单帧红外图像小目标检测达到更好的效果，弥补了工程人员缺少RPCA算法应用经验的不足。
- 红外小目标检测 /
- 鲁棒主成分分析 /
- 局部熵 /
- 预处理方案选择
Abstract: Based on the non-local similarity of images, the use of image block recombination to obtain low-rank block images is the basic method for applying robust principal component analysis (RPCA) for infrared small target detection involving single-frame image. This paper introduces the process of applying the RPCA algorithm in infrared small target detection involving single-frame images and analyzes the influence of various blocking methods under different image backgrounds. To address the difficulty of selecting the image block window and sliding step size under a complex background, a selection method based on the larger value of the minimum local entropy of the image block is proposed. The experimental results show that by calculating the block local entropy of the image, taking the larger value of the minimum local entropy as a reference, and selecting the RPCA algorithm preprocessing scheme, better results can be achieved in the detection of small targets in a single frame of infrared images. This addresses the lack of experience of engineering personnel with regard to the application of the RPCA algorithm.
- infrared small target detection /
- RPCA /
- local entropy /
- preprocessing scheme selection

HTML全文

图 1 单帧红外图像RPCA算法处理流程

Figure 1. Processing flow of single frame infrared image using RPCA algorithm

下载: 全尺寸图片幻灯片

图 2 不同背景图像及其块图像奇异值

Figure 2. Different background images and their block image singular value

下载: 全尺寸图片幻灯片

图 3 步长影响分析

Figure 3. Step length influence analysis

下载: 全尺寸图片幻灯片

图 4 窗口影响分析

Figure 4. Window size influence analysis

下载: 全尺寸图片幻灯片

图 5 最小局部熵随窗口变化

Figure 5. Minimum local entropy change with window size

下载: 全尺寸图片幻灯片

图 6 原始红外图像

Figure 6. Original infrared image

下载: 全尺寸图片幻灯片

图 7 背景抑制算法检测结果

Figure 7. Background suppression algorithm detection results

下载: 全尺寸图片幻灯片

图 8 随机选择(40×40, 4)预处理RPCA检测结果

Figure 8. Random selection(40×40, 4) preprocessed RPCA image detection results

下载: 全尺寸图片幻灯片

图 9 局部熵参考RPCA检测结果

Figure 9. Local entropy reference RPCA detection result

下载: 全尺寸图片幻灯片

表 1 背景抑制算法检测结果

Table 1. Background suppression algorithm detection results

Evaluation Index		Max Median	Tophat
SCRG	Fig.6(a) Fig.6(b) Fig.6(c)	29.7929 37.4023 23.3930	4.3761 6.9933 2.9932
BSF	Fig.6(a) Fig.6(b) Fig.6(c)	34.7819 43.4285 34.1074	26.7169 30.7303 27.1665

下载: 导出CSV

表 2 不同预处理RPCA检测结果

Table 2. RPCA detection results of different pretreatments

Evaluation Index		(12×12, 4)	(16×16, 4)	(20×20, 4)	(24×24, 4)
SCRG	Fig.6(a) Fig.6(b) Fig.6(c)	32.6057 18.0824 38.1424	41.6785 16.6169 40.2834	44.5846 13.4236 40.1180	36.5349 13.6778 39.5480
BSF	Fig.6(a) Fig.6(b) Fig.6(c)	40.3590 43.3497 54.9828	49.9203 42.2859 57.8680	52.9095 38.6833 58.0979	45.2896 37.4606 56.5966
		(30×30, 7)	(40×40, 4)	(50×50, 6)	(60×60, 4)
SCRG	Fig.6(a) Fig.6(b) Fig.6(c)	28.7508 14.5421 38.1673	23.0925 15.5531 39.0631	20.5614 15.8911 29.5542	20.2152 16.7226 35.7327
BSF	Fig.6(a) Fig.6(b) Fig.6(c)	35.4541 30.6043 51.0703	32.7816 33.3790 56.4791	27.3244 27.8227 43.2125	28.7974 29.6281 50.4376

下载: 导出CSV

表 3 各算法检测时间

Table 3. Detection time of each algorithm s

	Max Median	Tophat	RPCA (Local entropy reference)
Fig.6(a)	0.7942	0.8649	18.5329
Fig.6(b)	0.7895	0.6971	15.7956
Fig.6(c)	0.7969	0.5662	17.5864

下载: 导出CSV

参考文献(22)

[1]	Dehghani A, Pourmohammad A. Small target detection and tracking based on the background elimination and Kalman filter[C]//2015 The International Symposium on Artificial Intelligence and Signal Processing, 2015: 328-333.
[2]	YANG C, MA J, QI S, et al. Directional support value of Gaussian transformation for infrared small target detection[J]. Applied Optics, 2015, 54(9): 2255-2265. doi: 10.1364/AO.54.002255
[3]	WU S C, ZUO Z R. Small target detection in infrared images using deep convolutional neural networks[J]. Journal of Infrared and Millimeter Waves, 2019, 38(3): 371-380.
[4]	Deshpande S D, Er M H, Venkateswarlu R, et al. Max-mean and max-median filters for detection of small targets[C]//Signal and Data Processing of Small Targets 1999. International Society for Optics and Photonics, 1999, 3809: 74-83.
[5]	WANG P, TIAN J W, GAO C Q. Infrared small target detection using directional highpass filters based on ls-svm[J]. Electronics Letters, 2009, 45(3): 156-158. doi: 10.1049/el:20092206
[6]	ZHANG H, LEI Z H, DING XH. An improved method of median filter[J]. Journal of Image and Graphics, 2004, 9(4): 408-411. http://en.cnki.com.cn/Article_en/CJFDTOTAL-ZGTB200404004.htm
[7]	BAI X Z, ZHOU F G, XIE Y C, et al. Modified top-hat transformation based on contour structuring element to detect infrared small target[C]//IEEE Conference on Industrial Electronics and Applications, 2008, 3: 575-579.
[8]	Wright J, Allen Y, Ganesh A, et al. Robust face recognition via sparse representation[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009, 31(2): 210-227. doi: 10.1109/TPAMI.2008.79
[9]	史加荣, 郑秀云, 周水生. 矩阵补全算法研究进展[J]. 计算机科学, 2014, 41(4): 13-20. doi: 10.3969/j.issn.1002-137X.2014.04.003 SHI J R, ZHENG X Y, ZHOU S S. Research Progress in Matrix Completion Algorithms[J]. Computer Science, 2014, 41(4): 13-20. doi: 10.3969/j.issn.1002-137X.2014.04.003
[10]	陈川, 纪晓佳, 陈柘. 基于低秩矩阵恢复交通视频背景重建性能评价[J]. 计算机工程与设计, 2017, 38(5): 1301-1318. https://www.cnki.com.cn/Article/CJFDTOTAL-SJSJ201705034.htm CHEN C, JI X J, CHEN Z. Performance Evaluation of Traffic Video Background Reconstruction Based on Low-rank Matrix Recovery[J]. Computer Engineering and Design, 2017, 38(5): 1301-1318. https://www.cnki.com.cn/Article/CJFDTOTAL-SJSJ201705034.htm
[11]	杨敏, 安振英. 基于低秩矩阵恢复的视频背景建模[J]. 南京邮电大学学报, 2013, 33(2): 86-96. doi: 10.3969/j.issn.1673-5439.2013.02.015 YANG M, AN Z Y. Video Background Modeling Using Low-rank Matrix Recovery[J]. Journal of Nanjing University of Posts and Telecommunications, 2013, 33(2): 86-96. doi: 10.3969/j.issn.1673-5439.2013.02.015
[12]	GAO C Q, MENG D Y, YANG Y, et al. Infrared patch-image model for small target detection in a single image[J]. IEEE Transactions on Image Processing, 2013, 22(12): 4996-5009. doi: 10.1109/TIP.2013.2281420
[13]	EfrosA A, Leung T K. Texture synthesis by non-parametric sampling[C]//Proceedings of the seventh IEEE international conference on computer vision, 1999: 1033-1038.
[14]	马铭阳, 王德江, 孙翯, 等. 基于稳健主成分分析和多点恒虚警的红外弱小目标检测[J]. 光学学报, 2019, 39(8): 0810001-1-0810001-9. https://www.cnki.com.cn/Article/CJFDTOTAL-GXXB201908013.htm MA M Y, WANG D J, SUN H, et al. Infrared Dim-Small Target Detection Based on Robust Principal Component Analysis and Multi-Point Constant False Alarm[J]. Acta Optica Sinica, 2019, 39(8): 0810001-1-0810001-9. https://www.cnki.com.cn/Article/CJFDTOTAL-GXXB201908013.htm
[15]	樊俊良, 高永明, 吴止锾, 等. 基于RPCA的单帧红外小目标检测算法[J]. 兵器装备工程学报, 2018, 39(11): 147-151. doi: 10.11809/bqzbgcxb2018.11.032 FAN J L, GAO Y M, WU Z H, et al. Detection Algorithm of Single Frame Infrared Small Target Based on RPCA[J]. Journal of Ordnance Equipment Engineering, 2018, 39(11): 147-151. doi: 10.11809/bqzbgcxb2018.11.032
[16]	Wright J, Peng Y G, Ma Y, et al. Robust principal component analysis: exact recovery of corrupted low-rank matrices via convex optimization[J]. Advances in Neural Information Processing Systems, 2009, 87(4): 3-56.
[17]	Danelljan M, Khan FS, Felsberg M, et al, Adaptive color attributes for real-time visual tracking[C]//CVPR, 2014: 1090-1097.
[18]	史加荣, 郑秀云, 魏宗田, 等. 低秩矩阵恢复算法综述[J]. 计算机应用研究, 2013, 30(6): 1601-1605. doi: 10.3969/j.issn.1001-3695.2013.06.001 SHI J R, ZHENG X Y, WEI Z T, et al. Survey on Algorithms of Low-rank Matrix Recovery[J]. Application Research of Computers, 2013, 30(6): 1601-1605. doi: 10.3969/j.issn.1001-3695.2013.06.001
[19]	杜秀丽, 胡兴, 陈波, 等. 基于加权非局部相似性的视频压缩感知多假设重构算法[J]. 计算机科学, 2019, 46(1): 291-296. https://www.cnki.com.cn/Article/CJFDTOTAL-JSJA201901046.htm DU X L, HU X, CHEN B, et al. Multi-hypothesis Reconstruction Algorithm of DCVS Based on Weighted Non-local Similarity[J]. Computer Science, 2019, 46(1): 291-296. https://www.cnki.com.cn/Article/CJFDTOTAL-JSJA201901046.htm
[20]	陈湘凭, 王志成, 田金文. 基于局部梯度和局部熵的红外小目标融合检测[J]. 计算机与数字工程, 2006, 34(10): 1-4. doi: 10.3969/j.issn.1672-9722.2006.10.001 CHEN X P, WANG Z C, TIAN J W. Fusion Detection of Small Infrared Target Based on Local Entropy and Local Gradient Strength[J]. Computer & Digital Engineering, 2006, 34(10): 1-4. doi: 10.3969/j.issn.1672-9722.2006.10.001
[21]	毛羽忻, 杨俊强, 曲劲松, 等. 基于局部熵的点目标检测算法分析[J]. 火炮发射与控制学报, 2014, 35(3): 41-44. https://www.cnki.com.cn/Article/CJFDTOTAL-HPFS201403009.htm MAO Y X, YANG J Q, QU J S, et al. Analysis for Dot Target Detection Based on Local Entropy Algorithm[J]. Journal of Gun Launch & Control, 2014, 35(3): 41-44. https://www.cnki.com.cn/Article/CJFDTOTAL-HPFS201403009.htm
[22]	ZHANG Y X, LI L, XIN Y H. Infrared Small Target Detection Based on Adaptive Double-layer TDLMS Filter[J]. Acta Photonica Sinica, 2019, 48(9): 091001. http://en.cnki.com.cn/Article_en/CJFDTotal-GZXB201909023.htm