Dynamic Evaluation of Vehicle Camouflage Effect Based on Saliency Characterization in Complex Background

doi:10.12382/bgxb.2022.0181

Abstract

Abstract:

To meet the requirement of dynamic evaluation of the vehicle camouflage effect in complex landform background, a vehicle camouflage effect evaluation method based on superpixel segmentation and comprehensive saliency characterization was proposed. The method is based on the negative correlation between the saliency of the target in the background and the camouflage effect, the superpixel adjacency model is adopted, and the local saliency and the target saliency are introduced to comprehensively characterize the camouflage effect. Visual features of the vehicle and multi-scale images of the background are selected to construct multidimensional saliency matrix. Effective feature screening is carried out with the methods of complex correlation coefficient and variation coefficient. The entropy weight method is used to calculate the comprehensive saliency to evaluate the camouflage effect. Through verification and analysis of the numerical examples, the proposed method can evaluate the camouflage effect by integrating the complex landform background, vehicle camouflage state and classical salient visual features, and quantitatively analyze the saliency features and advantageous application environment of the vehicle camouflage state. This work aims to provide suggestions for vehicle camouflage design, state application and concealed area selection.

Key words: complex background, vehicle targets, saliency characterization, camouflage evaluation

WANG Dong, GAO Qinhe, HUANG Tong, LIU Zhihao, GAO Lei. Dynamic Evaluation of Vehicle Camouflage Effect Based on Saliency Characterization in Complex Background[J]. Acta Armamentarii, 2023, 44(6): 1620-1631.

Figures/Tables 14

References 31

[1]	陈鹏, 宋愿赟, 李文静, 等. 临近空间高速侦察与监视载荷技术研究综述[J]. 战术导弹技术, 2021(1): 7-12.
	CHEN P, SONG Y Y, LI W J, et al. Review of high speed reconnaissance and surveillance payload technology in near space[J]. Tactical Missile Technology, 2021(1): 7-12. (in Chinese)
[2]	XUE F, WU F, WANG J W, et al. Camouflage texture design based on camouflage performance evaluation[J]. Neurocomputing, 2018, 274(24):106-114. doi: 10.1016/j.neucom.2016.07.081 URL
[3]	贾其, 吕绪良, 吴超, 等. 人眼视觉注意机制用于伪装效果评价[J]. 应用科学学报, 2011, 29(3): 294-298.
	JIA Q, LÜ X L, WU C, et al. Evaluation of camouflage effectiveness using human visual attention mechanism[J]. Journal of Applied Sciences, 2011, 29(3): 294-298. (in Chinese)
[4]	邓建军, 高峰, 张琳. 基于减小观测距离的光学伪装效果等效检测[J]. 探测与控制学报, 2012, 34(3): 19-22.
	DENG J J, GAO F, ZHANG L. Equivalent test of optical camouflage effectiveness by reducing observation distance[J]. Journal of Detection & Control, 2012, 34(3): 19-22. (in Chinese)
[5]	崔宝生, 薛士强, 姬艳军, 等. 图像特征的伪装效果评估技术[J]. 红外与激光工程, 2010, 39(6): 1178-1183.
	CUI B S, XUE S Q, JI Y J, et al. Camouflage effectiveness evaluation based on image feature[J]. Infrared and Laser Engineering, 2010, 39(6): 1178-1183. (in Chinese)
[6]	王展, 颜云辉, 焦学勇. 基于灰色理论的迷彩伪装多指标综合评价[J]. 兵工学报, 2013, 34(10): 1250-1257. doi: 10. 3969/ j. issn. 1000-1093. 2013. 10. 009
	WANG Z, YAN Y H, JIAO X Y. Multi-index comprehensive evaluation of camouflage based on gray theory[J]. Acta Armamentarii, 2013, 34(10): 1250-1257. (in Chinese) doi: 10. 3969/ j. issn. 1000-1093. 2013. 10. 009
[7]	邓志鹏. 基于深度卷积神经网络的遥感图像目标检测方法研究[D]. 长沙: 国防科技大学, 2019.
	DENG Z P. Research on deep convolutional neural network based object detection methods in remote sensing images[D]. Changsha: National University of Defense University, 2019. (in Chinese)
[8]	胡建明, 乔凯, 智喜洋, 等. 天基观测条件下复杂环境对空中弱小目标可探测性的影响[J]. 红外与毫米波学报, 2019, 38(3): 351-357.
	HU J M, QIAO K, ZHI X Y, et al. Influence of complex environment on the detectability of weak and small aerial target under space-based observation mode[J]. Journal of Infrared and Millimeter Waves, 2019, 38(3): 351-357. (in Chinese)
[9]	宋明珠, 曲宏松, 金光. 含噪光学遥感图像海面弱小舰船目标检测[J]. 光学学报, 2017, 37(10):161-168.
	SONG M Z, QU H S, JIN G. Weak ship target detection of noisy optical remote sensing image on sea surface[J]. Acta Optica Sinica, 2017, 37(10):161-168. (in Chinese)
[10]	赵晓枫, 徐明扬, 王聃漂, 等. 基于改进SSD的特种车辆红外伪装检测方法[J]. 红外与激光工程, 2019, 48(11):116-125.
	ZHAO X F, XU M Y, WANG R P, et al. Infrared camouflage detection method for special vehiclesbased on improved SSD[J]. Infrared and Laser Engineering, 2019, 48(11):116-125. (in Chinese)
[11]	马世欣, 刘春桐, 李洪才, 等. 基于高光谱图像探测与感知的伪装效果评估方法[J]. 兵工学报, 2019, 40(7): 1485-1494. doi: 10.3969/j.issn.1000-1093.2019.07.019
	MA S X, LIU C T, LI H C, et al. Camouflage effect evaluation based on hyperspectral image detection and visual perception[J]. Acta Armamentarii, 2019, 40(7): 1485-1494. (in Chinese) doi: 10.3969/j.issn.1000-1093.2019.07.019
[12]	常沙, 宋彬, 黎鑫. 气象海洋作战仿真环境复杂性评估及分级方法[J]. 系统工程理论与实践, 2021, 41(7): 1868-1881. doi: 10.12011/SETP2019-2146
	CHANG S, SONG B, LI X. Complexity evaluation and classification method of meteorology and ocean operational simulation environments[J]. Systems Engineering-Theory & Practice, 2021, 41(7): 1868-1881. (in Chinese)
[13]	ACHANTA R, SHAJI A, SMITH K, et al. SLIC superpixels compared to state-of-the-art superpixel methods[J]. IEEE Transactions Pattern Analysis and Machine Intelligence, 2012, 34(11): 2274-2282. doi: 10.1109/TPAMI.2012.120 URL
[14]	ZHOU L, YANG Z H, ZHOU Z T, et al. Salient region detection using diffusion process on a two-layer sparse graph[J]. IEEE Transactions Image Processing, 2017, 26(12): 5882-5894. doi: 10.1109/TIP.2017.2738839 URL
[15]	吴加莹, 杨赛, 堵俊, 等. 自底向上的显著性目标检测研究综述[J]. 计算机科学, 2019, 46(3): 48-52. doi: 10.11896/j.issn.1002-137X.2019.03.006
	WU J Y, YANG S, DU J, et al. Review of bottom-up salient object detection[J]. Computer Science, 2019, 46(3): 48-52. (in Chinese)
[16]	ITTI L, KOCH C, NIEBUR E. A model of saliency-based visual attention for rapid scene analysis[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 1998, 20(11): 1254-1259. doi: 10.1109/34.730558 URL
[17]	QIAN D D, ZHOU Y F, WEI Y, et al. Graph laplacian based visual saliency detection[C]//Proceedings of the 2012 Fourth International Conference on Digital Home.Guangzhou,IEEE, 2012:201-205.
[18]	HOU X D, ZHANG L Q. Saliency detection: a spectral residual approach[C]//Proceedings of 2007 IEEE Conference on Computer Vision and Pattern Recognition. Minneapolis, MN, US: IEEE, 2007:1-8.
[19]	ACHANTA R, ESTRADA F, WILS P, et al. Salient region detection and segmentation[C]//Proceedings of International Conference on Computer Vision Systems.Berlin,Heidelberg, Germany:Springer, 2008:66-75.
[20]	ACHATA R, HEMAMI S, ESTRADA F, et al. Frequency-tuned salient region detection[C]//Proceedings of 2009 IEEE Conference on Computer Vision and Pattern Recognition. Miami,FL,US:IEEE, 2009:1597-1604.
[21]	GOFERMAN S, ZELNIK-MANOR L, TAL A. Context-aware saliency detection[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 34(10):1915-1926. doi: 10.1109/TPAMI.2011.272 URL
[22]	ACHATA R, SÜSSTRUNK S. Saliency detection using maximum symmetric surround[C]//Proceedings of 2010 IEEE International Conference on Image Processing. Hong Kong,China:IEEE, 2010:2653-2656.
[23]	PTRAZZI F, KRÄHENBÜHL P, PRITCH Y, et al. Saliency filters: contrast based filtering for salient region detection[C]//Proceedings of 2012 IEEE Conference on Computer Vision and Pattern Recognition. Providence, RI, US:IEEE, 2012:733-740.
[24]	LEE S M, XIN J H, WESTLAND S. Evaluation of image similarity by histogram intersection[J]. Wiley, 2005, 30(4): 265-274.
[25]	FLUSSER J, SUK T, ZITOV B. Moments and moment invariants in pattern recognition[M]. Chichester, UK: John Wiley & Sons, 2009.
[26]	YU H, LI M J, ZHANG H J, et al. Color texture moments for content-based image retrieval[C]//Proceedings of 2002 International Conference on Image Processing. Rochester, NY,US:IEEE, 2002:929-932.
[27]	EICHKITZ C G, AMTMANN J, SCHREILECHNER M G. Computation of grey level co-occurrence matrix (GLCM) attributes in single and multiple directions[J]. First Break, 2020, 38(3): 57-62.
[28]	WANG Z, BOVIK A C, SHEIKH H R, et al. Image quality assessment: from error visibility to structural similarity[J]. IEEE Transactions on Image Processing, 2004, 13(4): 600-612. doi: 10.1109/tip.2003.819861 pmid: 15376593
[29]	AHONEN T, HADID A, PIETIKAINEN M. Face description with local binary patterns: application to face recognition[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2006, 28(12): 2037-2041. pmid: 17108377
[30]	燕雪峰, 张德平, 黄晓冬, 等. 面向任务的体系效能评估[M]. 北京: 电子工业出版社, 2020.
	YAN X F, ZHANG D P, HUANG X D, et al. Task-oriented system effectiveness evaluation[M]. Beijing: Publishing House of Electronics Industry, 2020. (in Chinese)
[31]	XIA G S, BAI X, DING J, et al. DOTA: a large-scale dataset for object detection in aerial images[C]//Proceedings of 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition. Salt Lake City, UT,US:IEEE, 2018:3974-3983.

模型	图像空间	显著度表征
Itti等^[16] 模型	RGB	高斯金字塔法,对图像的亮度、颜色和4个方向的纹理进行“中心周边差分”操作
GBVS模型^[17]	RGB	用稳态平衡作为图像的亮度、颜色和4个方向的纹理的显著度衡量
SR模型^[18]	频域	对图像进行2维离散傅里叶变换,频域中log幅度谱的谱残差
Aho-Corasic 模型^[19]	CIE-Lab	Lab图像三个通道特征值与所有像素平均值的欧氏距离
FT模型^[20]	频域 CIE-Lab	频域滤波取反映图像整体信息的低频部分,计算Lab图像的欧氏距离
CA模型^[21]	CIE-Lab 空间距离	图像Lab三通道欧式距离和像素间的空间距离加权综合表征显著度
MSS模型^[22]	CIE-Lab	Lab图像三个通道与其固定邻域平均值的欧氏距离
SF模型^[23]	CIE-Lab 空间距离	综合分割图像的以空间距离为加权的区域颜色唯一性和以颜色相似度为加权的颜色空间分布

模型	图像空间	显著度表征
Itti等^[16] 模型	RGB	高斯金字塔法,对图像的亮度、颜色和4个方向的纹理进行“中心周边差分”操作
GBVS模型^[17]	RGB	用稳态平衡作为图像的亮度、颜色和4个方向的纹理的显著度衡量
SR模型^[18]	频域	对图像进行2维离散傅里叶变换,频域中log幅度谱的谱残差
Aho-Corasic 模型^[19]	CIE-Lab	Lab图像三个通道特征值与所有像素平均值的欧氏距离
FT模型^[20]	频域 CIE-Lab	频域滤波取反映图像整体信息的低频部分,计算Lab图像的欧氏距离
CA模型^[21]	CIE-Lab 空间距离	图像Lab三通道欧式距离和像素间的空间距离加权综合表征显著度
MSS模型^[22]	CIE-Lab	Lab图像三个通道与其固定邻域平均值的欧氏距离
SF模型^[23]	CIE-Lab 空间距离	综合分割图像的以空间距离为加权的区域颜色唯一性和以颜色相似度为加权的颜色空间分布

背景	伪装方案	有效特征
	1	Lab亮度、Hu矩、颜色矩、SR-LBP、SR-pHash
1	2	HSV颜色、 Hu矩、颜色矩、SR-LBP、SR-pHash
	3	Lab亮度、Hu矩、颜色矩、SR-LBP、SR-pHash
	1	Lab亮度、Hu矩、颜色矩、SR-LBP、SR-pHash
2	2	Lab亮度、Hu矩、颜色矩、SR-LBP、SR-pHash
	3	HSV颜色、 Hu矩、颜色矩、SR-LBP、SR-pHash

背景	伪装方案	有效特征
	1	Lab亮度、Hu矩、颜色矩、SR-LBP、SR-pHash
1	2	HSV颜色、 Hu矩、颜色矩、SR-LBP、SR-pHash
	3	Lab亮度、Hu矩、颜色矩、SR-LBP、SR-pHash
	1	Lab亮度、Hu矩、颜色矩、SR-LBP、SR-pHash
2	2	Lab亮度、Hu矩、颜色矩、SR-LBP、SR-pHash
	3	HSV颜色、 Hu矩、颜色矩、SR-LBP、SR-pHash

背景	伪装状态	最大变异系数	最大平均显著度
	1	SR-pHash	SR-LBP
1	2	SR-pHash	SR-LBP
	3	SR-pHash	颜色矩
	1	SR-pHash	SR-LBP
2	2	SR-pHash	颜色矩
	3	HSV颜色	SR-LBP