多参数最优重构水下偏振成像复原方法

doi:10.12382/bgxb.2022.0343

摘要/Abstract

摘要：

针对水下高浑浊度条件下成像存在清晰度低、对比度下降、图像质量低的难题,克服经典水下图像偏振复原方法需要进行无目标背景点选取的缺点,提出了一种多参数最优重构水下图像偏振复原方法。在经典水下成像物理模型基础上,将透射率细化为吸收系数与后向散射系数,通过引入Stokes矩阵计算水下图像偏振度;选取两个目标点,通过复原图像最优化,获取两个目标点反射率、吸收系数及后向散射系数的重构最优值,以去除水下图像后向散射并恢复吸收损失信号,实现水下图像偏振复原。通过两种无参客观评价指标与其他方法进行对比,不同浑浊度、不同目标下的实验结果表明,新方法能够实现水下降质图像复原,特别是在高浑浊度条件下更为有效。新方法有望应用于水下航行器中的光学清晰成像,便于后续目标检测。

关键词: 水下图像复原, 偏振成像, 最优重构, 多参数

Abstract:

Underwater imaging in high turbidity conditions often suffers from issues such as low resolution, reduced contrast, and overall poor image quality. Classical methods for underwater image polarization restoration require the selection of a background point without any target, making them inapplicable in certain scenarios. In order to solve this problem, the method of underwater polarization image restoration based on optimal multi-parameter reconstruction is proposed. Based on the classical underwater imaging physical model, the transmittance is divided into absorption and backscattering coefficients. By calculating the polarization degree of the underwater image using the Stokes vector, two target points are selected. The optimal reconstruction values of reflectivity, absorption coefficient and backscattering coefficient of two target points are obtained by optimizing the restored image. By using the optimized parameters to remove the backscattered light and recover the signal light lost due to absorption from the underwater image, the polarization restoration of the underwater image is realized. Two no-reference image quality assessment indexes are employed as quantitative indexes. Compared with other methods, experiments based on different turbidity and different targets show that this method can effectively restore the degraded underwater image, especially in the case of high turbidity. This method is expected to be applied to enhance optical imaging clarity of underwater vehicles and facilitate subsequent target detection.

Key words: underwater image restoration, polarization image, optimal reconstruction, multi-parameters

陈雄锋, 阮驰. 多参数最优重构水下偏振成像复原方法[J]. 兵工学报, 2023, 44(7): 2122-2131.

CHEN Xiongfeng, RUAN Chi. Underwater Polarization Image Restoration Method Using Optimal Multi-Parameters Reconstruction[J]. Acta Armamentarii, 2023, 44(7): 2122-2131.

图/表 11

图1 改进后的水下成像物理模型示意图

Fig.1 Improved physical model for underwater imaging

图2 基于多参数最优重构的水下偏振成像复原方法流程

Fig.2 Flow chart of underwater polarization image restoration method based on optimal multi-parameter reconstruction

图3 不同浑浊度下光强度原图

Fig.3 Light intensities under different turbidity conditions

表1 不同浑浊度下3种方法的处理结果

Table 1 Results of three methods under different turbidity levels

图4 不同浑浊度下对比度(上)与EME(下)评价结果

Fig.4 Results of contrast (top) and EME (bottom) under different turbidity levels

表2 不同浑浊度下处理结果(光学分辨率板反片)

Table 2 Results under differentturbidity levels (reverse side of the resolution test plate)

表3 不同浑浊度下处理结果(钥匙扣)

Table 3 Results under different turbidity levels (key chain)

图5 不同浑浊度下对比度(上)与EME(下)评价结果(光学分辨率板反片)

Fig.5 Results of contrast (top) and EME (bottom) under different turbidity levels (reverse side of the resolution test plate)

图6 不同浑浊度下对比度(上)与EME(下)评价结果(钥匙扣)

Fig.6 Results of contrast (top) and EME (bottom) under different turbidity levels (key chain)

图7 不同距离误差下复原结果图(8mL浑浊度下的钥匙扣)

Fig.7 Results under different distance errors(key chain at 8mL turbidity)

表4 不同距离误差下的对比度

Table 4 Results of contrast degrees with different distance errors

距离/cm	原图	10	20	30	40
对比度	0.81	20.74	23.05	24.26	25.71

参考文献 22

[1]	全向前, 陈祥子, 全永前, 等. 深海光学照明与成像系统分析及进展[J]. 中国光学, 2018, 11(2):153-165.
	QUAN X Q, CHEN X Z, QUAN Y Q, et al. Analysis and research progress of deep-sea optical illumination and imaging system[J]. Chinese Optics, 2018, 11(2):153-165. (in Chinese) doi: 10.3788/co. URL
[2]	谭克, 邹丽娜. 正态分布转换算法在声纳图像处理中的应用[J]. 兵工学报, 2016, 37(6): 1052-1057. doi: 10.3969/j.issn.1000-1093.2016.06.012
	TAN K, ZOU L N. The Application of normal distribution transformation algorithm in sonar image processing[J]. Acta Armamentarii, 2016, 37(6): 1052-1057. (in Chinese) doi: 10.3969/j.issn.1000-1093.2016.06.012
[3]	LEE H S, SANG W M, EOM I K. Underwater image enhancement using successive color correction and superpixel dark channel prior[J]. Symmetry, 2020, 12(8):1220. doi: 10.3390/sym12081220 URL
[4]	LUO W L, DUAN S Q, ZHENG J W. Underwater image restoration and enhancement based on a fusion algorithm with color balance,contrast optimization and histogram stretching[J]. IEEE Access, 2021, 99:1-1.
[5]	赵永强, 戴慧敏, 申凌皓, 等. 水下偏振清晰成像方法综述[J]. 红外与激光工程, 2020, 49(6):20190574-1.
	ZHAO Y Q, DAI H M, SHEN L H, et al. Review of underwater polarization clear imaging methods[J]. Infrared and Laser Engineering, 2020, 49(6):20190574-1. (in Chinese)
[6]	LIN J Q, YU M, XU H Y, et al. Underwater image restoration based on light attenuation prior and background light fusion[J]. Laser & Optoelectronics Progress, 2021, 58(8):0810013.
[7]	ULUTAS G, USTUBIOGLU B. Underwater image enhancement using contrast limited adaptive histogram equalization and layered difference representation[J]. Multimedia Tools and Applications, 2021(20):15067-15091.
[8]	SOMASEKAR M, MURUGAN S S. Fusion-based approach for quality enhancement of underwater images[J]. Journal of Environmental Protection and Ecology, 2021, 22(4):1676-1687.
[9]	SINGH B, MISHRA R S, GOUR P. Analysis of contrast enhancement techniques for underwater image[J]. International Journal of Computer Technology and Electronics Engineering, 2011, 1(2):190-194.
[10]	GARG D, GARGN K, KUMAR M. Underwater image enhancement using blending of CLAHE and percentile methodologies[J]. Multimedia Tools and Applications, 2018, 77(20):26545-61. doi: 10.1007/s11042-018-5878-8
[11]	MCGLAMERY B L. A computer model for underwater camera systems[J]. Ocean Optics VI, 1979, Monterey, United States. Proceedings of the SPIE, 1980, 208(208):221-231.
[12]	HE K M, SUN J, TANG X O. Single image haze removal using dark channel prior[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2011, 33(12):2341-2353. doi: 10.1109/TPAMI.2010.168 pmid: 20820075
[13]	GALDRAN A, VAZQUEZ-CORRAL J, PARDO D, et al. A variational framework for single image dehazing[C]// Proceedings of European Conference on Computer Vision.Berlin,Germany:Springer, 2014.
[14]	LI C Y, GUO J C, CONG R M, et al. Underwater image enhancement by dehazing with minimum information loss and histogram distribution prior[J]. IEEE Transactions on Image Processing, 2019, 26(12):5664-5677.
[15]	SCHECHNER Y Y, NARASIMHAN G, NAYAR S K. Instant dehazing of images using polarization[C]//Proceedings of the 2001 IEEE Computer Society Conference on Computer Vision and Pattern Recognition,CVPR 2001, Kauai, HI, US:IEEE, 2001.
[16]	SCHECHNER Y Y, KARPEL N. Clear underwater vision[C]//Proceedings of the 2004 IEEE Computer Society Conference on Computer Vision and Pattern Recognition.CVPR 2004, Washington, DC, US:IEEE, 2004.
[17]	SCHECHNER Y Y, KARPEL N. Recovery of underwater visibility and structure by polarization analysis[J]. IEEE Journal of Oceanic Engineering, 2006, 30(3):570-587. doi: 10.1109/JOE.2005.850871 URL
[18]	LIU F, ZHANG S C, HAN P L, et al. Depolarization index from Mueller matrix descatters imaging in turbid water[J]. Chinese Optics Letters, 2022, 20(2): 022601. doi: 10.3788/COL URL
[19]	HUANG B J, LIU T G, HU H F, et al. Underwater image recovery considering polarization effects of objects[J]. Optics Express, 2016, 24(9): 9826-9838. doi: 10.1364/OE.24.009826 pmid: 27137596
[20]	WU H D, ZHAO M, LI F Q, et al. Underwater polarization-based single pixel imaging[J]. Journal of the Society for Information Display, 2020, 28(2):157-163. doi: 10.1002/jsid.838
[21]	BARUKČIČ I, SUN Z P, LI F B, et al. Underwater image enhancement algorithm based on dark channel prior and underwater imaging model[J]. MATEC Web of Conferences, 2021, 336,06033. doi: 10.1051/matecconf/202133606033 URL
[22]	李荣华, 唐智超, 朴俊峰, 等. 偏振参数最优重构的水下降质图像清晰化方法[J]. 红外与激光工程, 2021, 50(6):20200426-1.
	LI R H, TANG Z C, PIAO J F, et al. Underwater degraded image-sharpening method based on optimal polarization parameter reconstruction[J]. Infrared and Laser Engineering, 2021, 50(6):20200426-1. (in Chinese)