全偏振参量低秩稀疏分解伪彩色图像融合

doi:10.3969/j.issn.1001-506X.2020.11.06

系统工程与电子技术 ›› 2020, Vol. 42 ›› Issue (11): 2450-2460.doi: 10.3969/j.issn.1001-506X.2020.11.06

全偏振参量低秩稀疏分解伪彩色图像融合

徐国明^1,^2,³(), 袁宏武^2,³(), 薛模根³(), 王峰³()

1. 安徽大学互联网学院, 安徽合肥 230039
2. 安徽新华学院信息工程学院, 安徽合肥 230088
3. 陆军炮兵防空兵学院偏振光成像探测技术安徽省重点实验室, 安徽合肥 230031

收稿日期:2020-03-29 出版日期:2020-11-01 发布日期:2020-11-05
作者简介:徐国明(1979-),男,副教授,博士,主要研究方向为稀疏表示、计算机视觉、图像处理。E-mail:xgm121@163.com|袁宏武(1979-),男,副教授,博士,主要研究方向为偏振成像探测、机器学习。E-mail:931030372@qq.com|薛模根(1964-),男,教授,博士,主要研究方向为光电防御、图像处理。E-mail:xuemogen@126.com|王峰(1972-)，男，教授，博士，主要研究方向为新型光电成像探测技术。E-mail:wfissky@sina.com
基金资助:
国家自然科学基金(61379105);安徽省自然科学基金(1608085MF140);安徽省自然科学基金(1908085MF208);中国博士后科学基金(2016M592961);陆军装备部十三五预研子课题(GFZX0403260204);安徽省高校自然科学研究重点项目(KJ2018A0587);安徽省高校自然科学研究重点项目(KJ2019A0906)

Full polarization parameters low-rank and sparse factorization for pseudo color image fusion

Guoming XU^1,^2,³(), Hongwu YUAN^2,³(), Mogen XUE³(), Feng WANG³()

1. College of Internet, Anhui University, Hefei 230039, China
2. Information Engineering College, Anhui Xinhua University, Hefei 230088, China
3. Anhui Province Key Laboratory of Polarized Imaging Detecting

Received:2020-03-29 Online:2020-11-01 Published:2020-11-05

摘要/Abstract

摘要：

偏振图像伪彩色融合对提高视觉感知和目标判读具有重要意义,利用空间调制型全偏振参量矩阵的低秩和稀疏特性,提出基于贝叶斯概率鲁棒性矩阵分解融合方法。首先,根据偏振调制和解析算法构造偏振参量矩阵,同时合成强度图像;其次,对参量矩阵进行基于改进的贝叶斯概率参量矩阵分解,降低背景噪声和亮度变化等干扰,分别获得参量图像的稀疏和低秩成分;然后利用方差、清晰度和信息熵进行模糊积分,获得显著性参量图像,与合成强度图像一起进行像素级增强;最后,经直方图规定化和IHS颜色映射,得到伪彩色融合结果。实验选择多种材质与目标的仿真和实测数据进行验证,通过主观视觉效果和客观指标比较,验证了其有效性。

关键词: 偏振成像, 图像融合, 低秩稀疏, 鲁棒性矩阵分解

Abstract:

The polarization image pseudo color fusion is of great significance to improve visual perception and object interpretation. Based on the low-rank and sparse prior of the spatially modulated full-polarization parameter matrix, a fusion method is proposed via Bayesian probability robust matrix factorization. Firstly, the polarization parameter matrix is constructed according to the polarization modulation and analytic algorithm, and the intensity image is synthesized at the same time. Secondly, the sparse and low-rank components of the parameter images are obtained via improved Bayesian robust matrix factorization (BRMF). The influence of background noise and brightness change is reduced via BRMF. Then, the saliency parameter image is obtained by using the Choquet fuzzy integral of variance, definition and entropy. The saliency parameter image is carried out together with the synthesized intensity image to enforce pixel level enhancement. Finally, the pseudo color fusion image is obtained by the processing of histogram normalization and IHS color mapping. The experiment is performed with simulation and measurement images of different materials and objects. The subjective visual results and objective evaluation verify the effectiveness of the proposed method.

Key words: polarization imaging, image fusion, low-rank and sparse, robust matrix factorization

中图分类号:

徐国明, 袁宏武, 薛模根, 王峰. 全偏振参量低秩稀疏分解伪彩色图像融合[J]. 系统工程与电子技术, 2020, 42(11): 2450-2460.

Guoming XU, Hongwu YUAN, Mogen XUE, Feng WANG. Full polarization parameters low-rank and sparse factorization for pseudo color image fusion[J]. Systems Engineering and Electronics, 2020, 42(11): 2450-2460.

图/表 10

图1

图2

图3

图4

表1

图5

图6

图7

图8

表2

参考文献 23

1	王小龙, 王峰, 刘晓, 等. 荒漠背景下典型伪装目标的高光谱偏振特性[J]. 激光与光电子学进展, 2018, 55 (5): 051101.
	WANG X L , WANG F , LIU X , et al. Hyperspectral polarization characteristics of typical camouflage target under desert background[J]. Laser & Optoelectronics Progress, 2018, 55 (5): 051101.
2	ZHAO Y , YI C , KONG S G , et al. Multi-band polarization imaging and applications[M]. Berlin: Springer Publishing Company, Incorporated, 2016: 47- 71.
3	SHUMYATSKY P , MILIONE G , ALFANO R R . Optical memory effect from polarized Laguerre-Gaussian light beam in light-scattering turbid media[J]. Optics Communications, 2014, 321, 116- 123. doi: 10.1016/j.optcom.2014.01.039
4	张家民, 时东锋, 黄见, 等. 图像融合在偏振关联成像中的应用[J]. 红外与激光工程, 2018, 47 (12): 373- 379.
	ZHANG J M , SHI D F , HUANG J , et al. Application of image fusion in polarization correlated imaging[J]. Infrared and Laser Engineering, 2018, 47 (12): 373- 379.
5	ZHANG L , YANG F B , JI L N , et al. Multiple-algorithm parallel fusion of infrared polarization and intensity images based on algorithmic complementarity and synergy[J]. Journal of Electronic Imaging, 2018, 27 (1) doi: 10.1117/1.JEI.27.1.013029
6	TOET A . Natural colour mapping for multiband night vision imagery[J]. Information Fusion, 2003, 4 (3): 155- 166. doi: 10.1016/S1566-2535(03)00038-1
7	WOLFF L B . Polarization vision: a new sensory approach to image understanding[J]. Image and Vision Computing, 1997, 15 (2): 81- 93. doi: 10.1016/S0262-8856(96)01123-7
8	宫剑, 吕俊伟, 刘亮. 基于颜色空间融合与上下文显著性的红外偏振图像目标增强[J]. 光学学报, 2019, 39 (10): 23- 32.
	GONG J , LYU J W , LIU L . Target enhancement of infrared polarization image based on color space fusion and context-aware saliency[J]. Acta Optica Sinica, 2019, 39 (10): 23- 32.
9	LEE D D , SEUNG H S . Learning the parts of objects by non-negative matrix factorization[J]. Nature, 1999, 401 (6755): 788- 791. doi: 10.1038/44565
10	周浦城, 韩裕生, 薛模根, 等. 基于非负矩阵分解和IHS颜色模型的偏振图像融合方法[J]. 光子学报, 2010, 39 (9): 1682- 1687.
	ZHOU P C , HAN Y S , XUE M G , et al. Polarization image fusion approach based on non-negative matrix factorization and ihs color model[J]. Acta Photonica Sinica, 2010, 39 (9): 1682- 1687.
11	SALAKHUTDINOV R , MNIH A . Probabilistic matrix factorization[J]. Advances in Neural Information Processing Systems, 2008, 20, 1257- 1264.
12	常侃, 张智勇, 陈诚, 等. 采用低秩与加权稀疏分解的视频前景检测算法[J]. 电子学报, 2017, 45 (9): 2272- 2280.
	CHANG K , ZHANG Z Y , CHEN C , et al. Video foreground detection by low-rank and reweighted sparse decomposition[J]. Acta Electronica Sinica, 2017, 45 (9): 2272- 2280.
13	TATSUYA Y, KAZUYA K, MUNEYUKI S, et al. Dynamic PET image reconstruction using nonnegative matrix factorization incorporated with deep image prior[C]//Proc.of the IEEE International Conference on Computer Vision, 2019: 3126-3135.
14	宗静静, 邱天爽. 稀疏表示中稀疏系数的l₁范数的特性分析[J]. 系统工程与电子技术, 2019, 41 (12): 2692- 2696.
	ZONG J J , QIU T S . Characteristic analysis for the l₁ norm of sparse coefficients in sparse representation[J]. Systems Engineering and Electronics, 2019, 41 (12): 2692- 2696.
15	PENG Y , GANESH A , WRIGHT J , et al. RASL: robust alignment by sparse and low-rank decomposition for linearly correlated images[J]. IEEE Trans.on Pattern Analysis & Machine Intelligence(TPAMI), 2012, 34 (11): 2233- 2246.
16	ZHANG Z , GANESH A , LIANG X , et al. TILT: Transform invariant low-rank textures[J]. International Journal of Computer Vision, 2012, 99 (1): 1- 24.
17	WRIGHT J, PENG Y G, MA Y. Robust principal component analysis: exact recovery of corrupted low-rank matrices by convex optimization[C]//Proc.of the 23rd Annual Conference on Neural Information Processing Systems, 2009: 2080-2088.
18	SALAKHUTDINOV R, MNIH A. Bayesian probabilistic matrix factorization using Markov chain Monte Carlo[C]//Proc.of the 25th International Conference Machine Learning, 2008: 880-887.
19	WANG N Y, YEUNG D Y. Bayesian robust matrix factorization for image and video processing[C]//Proc.of the IEEE International Conference on Computer Vision, 2013: 1785-1792.
20	WANG N, YAO T, WANG J, et al. A probabilistic approach to robust matrix factorization[C]//Proc.of the European Conference on Computer Vision, 2012: 126-139.
21	SMITH, RAY A. Color gamut transform pairs[C]//Proc.of the 5th Annual Conference on Computer Graphics and Interactive Techniques, 1978: 12-19.
22	安影, 范训礼, 陈莉, 等. 结合FABEMD和改进的显著性检测的图像融合[J]. 系统工程与电子技术, 2020, 42 (2): 292- 300.
	AN Y , FAN X L , CHEN L , et al. Image fusion combining FABEMD with improved saliency detection[J]. Systems Engineering and Electronics, 2020, 42 (2): 292- 300.
23	ZHAO Q , MENG D Y , XU Z B , et al. L1-norm low-rank matrix factorization by variational Bayesian method[J]. IEEE Trans.on Neural Networks and Learning Systems, 2015, 26 (4): 825- 839. doi: 10.1109/TNNLS.2014.2387376

H取值范围	变换方法
[0°, 120°]	${B = I(1- S)}$
	${R = I\left({1 + \frac{{S\cos H}}{{\cos \left({{{60}^\circ } - H} \right)}}} \right)}$
	${G = 3I- (B + R)}$
[120°, 240°]	${R = I(1- S)}$
	${G = I\left({1 + \frac{{S\cos \left({H - {{120}^\circ }} \right)}}{{\cos \left({{{180}^\circ } - H} \right)}}} \right)}$
	${B = 3I- (G + R)}$
[240°, 360°]	${G = I(1- S)}$
	${B = I\left({1 + \frac{{Scos\left({H - {{240}^\circ }} \right)}}{{\cos \left({{{300}^\circ } - H} \right)}}} \right)}$
	${R = 3I- (B + G)}$

目标	Rce			Cor			Con
目标	NMF	SNMF	Proposed	NMF	SNMF	Proposed	NMF	SNMF	Proposed
fabric	9.32	9.45	8.42	0.39	0.17	0.85	25.63	34.85	79.58
mask	12.10	7.30	9.80	0.09	0.35	0.66	20.20	43.88	39.13
tank	12.97	12.72	10.94	-0.11	-0.22	0.77	83.42	70.96	90.98
aircraft	11.76	10.74	9.32	0.44	-0.23	0.19	41.79	35.41	93.67
ship	12.15	11.20	9.55	0.31	0.29	0.41	51.30	74.01	68.61
平均	11.66	10.28	9.61	0.22	0.07	0.58	44.47	51.82	74.39

[1]	王健, 何自豪, 刘洁, 杨珂. 基于梯度域导向滤波器和改进PCNN的图像融合算法[J]. 系统工程与电子技术, 2022, 44(8): 2381-2392.
[2]	童莹萍, 全英汇, 冯伟, 邢孟道. 基于空谱信息协同与Gram-Schmidt变换的多源遥感图像融合方法[J]. 系统工程与电子技术, 2022, 44(7): 2074-2083.
[3]	杨真真, 乐俊, 杨永鹏, 范露. 非凸运动辅助低秩稀疏分解目标检测算法[J]. 系统工程与电子技术, 2020, 42(6): 1218-1225.
[4]	安影, 范训礼, 陈莉, 刘佩. 结合FABEMD和改进的显著性检测的图像融合[J]. 系统工程与电子技术, 2020, 42(2): 292-300.
[5]	宗静静, 邱天爽. 稀疏表示中稀疏系数的 l-1 范数的特性分析[J]. 系统工程与电子技术, 2019, 41(12): 2692-2696.
[6]	刘斌, 付忠旺. 基于四通道不可分提升小波的多聚焦图像融合[J]. 系统工程与电子技术, 2018, 40(2): 463-471.
[7]	常莉红, 冯象初, 张瑞. 四元数小波变换联合稀疏表示的图像融合[J]. 系统工程与电子技术, 2017, 39(7): 1633-1639.
[8]	赵蓝飞, 王爱丽, 王博, 吕鑫淼. 基于子图融合技术的图像增强算法[J]. 系统工程与电子技术, 2017, 39(12): 2840-2848.
[9]	胡晓伟, 童宁宁, 何兴宇, 雷腾, 王宇晨. 基于窄带雷达组网的高速自旋目标成像[J]. 系统工程与电子技术, 2016, 38(4): 792-798.
[10]	周志强，汪渤，李立广，李笋. 基于双边与高斯滤波混合分解的图像融合方法[J]. 系统工程与电子技术, 2016, 38(1): 8-13.
[11]	刘斌, 刘维杰, 魏艳萍. 6通道MSVD构造及其在多聚焦图像融合中的应用[J]. 系统工程与电子技术, 2015, 37(9): 2191-2197.
[12]	邢雅琼, 王晓丹, 梁兵杰, 秦卓. 基于非下采样剪切波变换域方向信息测度的多聚焦图像融合方法[J]. 系统工程与电子技术, 2015, 37(1): 191-198.
[13]	郭明，王书满. 基于区域和方向方差加权信息熵的图像融合[J]. Journal of Systems Engineering and Electronics, 2013, 35(4): 720-724.
[14]	邢雅琼，王晓丹，毕　凯，雷　蕾. 基于NSCT和ICA的红外和可见光图像融合方法[J]. 系统工程与电子技术, 2013, 35(11): 2251-2257.
[15]	王晓文，赵宗贵，汤磊. 一种新的红外与可见光图像融合评价方法[J]. Journal of Systems Engineering and Electronics, 2012, 34(5): 871-875.

全偏振参量低秩稀疏分解伪彩色图像融合

Full polarization parameters low-rank and sparse factorization for pseudo color image fusion

RichHTML

PDF (PC)

可视化

摘要/Abstract

引用本文

使用本文

图/表 10

参考文献 23

相关文章 15

编辑推荐

Metrics

本文评价