激光水下成像噪声分析及图像处理方法研究
|
摘要
水下激光成像技术在水下探测等海洋开发中具有重要的实用价值,但由于水介质的特殊性,使其所成图像易受到噪声污染而造成分辨率下降。即使采用了同步扫描和距离选通等技术,获得的成像质量仍然不好,图像存在由后向散射形成的散斑噪声和其他一些高斯噪声的影响。为了进一步改善成像效果,就需要利用数字图像处理技术对水下成像进行后续加工,水下图像处理结果的好坏直接影响到后面的特征提取和目标识别等技术的成败。
本文首先介绍了近年发展起来的几种蓝绿激光水下目标探测技术,即距离选通、同步扫描、条纹管3D成像技术和偏振激光成像等,并介绍了相关的探测器件性能参数。
其次,研究激光在水中传输的衰减特性中,分析了水体对光的“吸收”和“散射”效应,特别是后向散射的影响,给水下激光成像带来大量的背景噪声,是水下成像高噪声和低对比度的主要成因。特别针对后向散射进行研究,建立重叠系数分析模型,根据典型系统参数计算并绘制了系统的重叠系数和后向散射光随时间的变化曲线。为了验证图像诸多噪声类型中,哪种噪声占主要因素,通过分析分别受到散斑噪声和高斯噪声污染图像的概率曲线及灰度均值方差比等统计特性,找出两种噪声分别作用下的差别,以此作为判断标准,对激光水下图像进行分析比对,可知图像所受噪声的主要成分为散斑噪声,从而为后续图像处理技术提供了依据。
最后,在对水下图像进行滤波处理过程中,应用MATLAB软件分析比较了几种常用滤波方法的图像增强效果,验证了中值滤波在图像处理中的优势所在。通过比较各种算法处理图像的散斑指数、峰值信噪比(PSNR)和归一化均方误差(NMSE),比较了它们的散斑抑制能力
The technology of underwater laser imaging has very important practical value at ocean exploitation, but its imagery is usually affected by speckle noise. Although synchronous scanning technology and range-gated technology are mostly applied, it can hardly guarantee the perfect processing results. A large number of background noises are brought to underwater laser image for the reasons of absorption and scattering of water, which is also the cause of noises and low contrast of the underwater images. In order to improve the effect of the images, digital image processing techniques are introduced.
This paper introduces several technology of green-blue laser underwater target detection developed in recent years such as range gated imaging system, synchronous scanning, polarized laser imaging, streak tube imaging lidar. The respective imaging principle, characteristics and update status are presented.
To analysis the back scatter of a imaging system, the optical back scatter of water arriving at a detector is calculated, a model of optical back scatter of the underwater imaging system is given. The theoretical calculations are programmed, and figures of overlap coefficient and optical back scatter plotted based on typical system parameters. Which noise on the dominating status is very important for filtering algorithm, so, the statistic characteristics such as probability curve and gray mean variance retioetc. In the imagery disturbed by speckle and some other gaussian noises are analyzed firstly, and then the different statistic characteristics affected by the two kinds of noises respectively are found. Based on these characteristics, the images of underwater laser imaging system are discussed, and then the results show that the most imagery noise is speckle.
This paper displays the powerful ability of MATLAB in image processing with underwater laser image processing. Median filtering is a non-linear image processing method that can preserve image edge while removing noise. In light of computing the speckle indexes, PSNR and NMSE of the processed images, their ability of suppressing speckle noise is compared.
本文首先介绍了近年发展起来的几种蓝绿激光水下目标探测技术,即距离选通、同步扫描、条纹管3D成像技术和偏振激光成像等,并介绍了相关的探测器件性能参数。
其次,研究激光在水中传输的衰减特性中,分析了水体对光的“吸收”和“散射”效应,特别是后向散射的影响,给水下激光成像带来大量的背景噪声,是水下成像高噪声和低对比度的主要成因。特别针对后向散射进行研究,建立重叠系数分析模型,根据典型系统参数计算并绘制了系统的重叠系数和后向散射光随时间的变化曲线。为了验证图像诸多噪声类型中,哪种噪声占主要因素,通过分析分别受到散斑噪声和高斯噪声污染图像的概率曲线及灰度均值方差比等统计特性,找出两种噪声分别作用下的差别,以此作为判断标准,对激光水下图像进行分析比对,可知图像所受噪声的主要成分为散斑噪声,从而为后续图像处理技术提供了依据。
最后,在对水下图像进行滤波处理过程中,应用MATLAB软件分析比较了几种常用滤波方法的图像增强效果,验证了中值滤波在图像处理中的优势所在。通过比较各种算法处理图像的散斑指数、峰值信噪比(PSNR)和归一化均方误差(NMSE),比较了它们的散斑抑制能力
The technology of underwater laser imaging has very important practical value at ocean exploitation, but its imagery is usually affected by speckle noise. Although synchronous scanning technology and range-gated technology are mostly applied, it can hardly guarantee the perfect processing results. A large number of background noises are brought to underwater laser image for the reasons of absorption and scattering of water, which is also the cause of noises and low contrast of the underwater images. In order to improve the effect of the images, digital image processing techniques are introduced.
This paper introduces several technology of green-blue laser underwater target detection developed in recent years such as range gated imaging system, synchronous scanning, polarized laser imaging, streak tube imaging lidar. The respective imaging principle, characteristics and update status are presented.
To analysis the back scatter of a imaging system, the optical back scatter of water arriving at a detector is calculated, a model of optical back scatter of the underwater imaging system is given. The theoretical calculations are programmed, and figures of overlap coefficient and optical back scatter plotted based on typical system parameters. Which noise on the dominating status is very important for filtering algorithm, so, the statistic characteristics such as probability curve and gray mean variance retioetc. In the imagery disturbed by speckle and some other gaussian noises are analyzed firstly, and then the different statistic characteristics affected by the two kinds of noises respectively are found. Based on these characteristics, the images of underwater laser imaging system are discussed, and then the results show that the most imagery noise is speckle.
This paper displays the powerful ability of MATLAB in image processing with underwater laser image processing. Median filtering is a non-linear image processing method that can preserve image edge while removing noise. In light of computing the speckle indexes, PSNR and NMSE of the processed images, their ability of suppressing speckle noise is compared.
引文
1.郑贵斌.海洋资源开发集成战略与海洋经济可持续创新发展.理论学刊,2006,6(6):37~39
2. L.R.Marshall. Rlue-green lasers plumb the mysteries of the deep. Laser Focus World,1993,4(4):185~197
3.孔捷,张保民.激光水下成像技术及其进展.光电子技术.2006,26(2):130~132
4.杨华勇,梁永辉.机载蓝绿激光水下目标探测技术的现状及前景.光机电信息,2003,12(1):6~10
5.朱仁慈.适合水下检查的激光成像.国外地质勘探技术,1993,4(2):14~23
6.李荣福,崔佳华,李桂娟.蓝绿脉冲激光用于海深及水下目标探测.声学与电子工程,1990,6(1):34~40
7.昌彦君,彭复员.水下激光成像的实验研究.实验室研究与探索.2009,28(3):19~20
8.李哲,邓甲昊,周卫平.水下激光探测技术及其进展.舰船电子工程.2008,28(12):8~11
9. A.GORDON. Use of laser scanning system on mobile underwater platforms. IEEE,1992,7803:202~205
10.刘雪明,谭志飞,胡正荣.水下成像系统及主要参数的理论计算.光电子激光,1999,9(1):71~72
11.刘雪明,张明德,孙小茜.水下激光成像系统的MTF分析.中国激光,1999,26(4):332~336
12. L.E.Mertens. Use of point spread and beam spread functions for analysis of imaging systems in water. J Opt. Soc. Am,1977,67(8):1105~1117
13.赵岩,翟百臣,王建立.激光距离选通成像同步控制系统的设计与实现.红外与激光工程.2005,34(5):527~529
14.孙健,周亚民.距离选通激光水下成像系统研究.舰船电子工程,2009,29(2):24~26
15.柏连发,张毅,陈钱.距离选通成像实现过程中若干问题的探讨.红外与激光工程,2009,38(1):58~61
16.昌彦君,彭复员.水下激光成像的实验研究.实验室研究与探索,2009,28(3):19~21
17.王祥科,张辉,郑志强.距离选通技术对后向散射抑制作用的理论分析.强激光与粒子束.2007,19(12):1998~2000
18.朱耘.激光水下三维成像技术及进展.应用光学,1999,20(4):22~26
19.李思宁,刘金波,李慧子.大视场情况下激光雷达距离方程的修正.中国激光,2006,33(suppl):108~110
20. D.Anthony, H.Gleckler, A.Gelbart, et al. Multispectral and hyperspectral 3D imaging lidar based upon the multiple slit streak tube imaging lidar. Proc. SPIE,2001,4377(8):328~335
21. D.Anthony, H.Gleckler, A.Gelbart. Three-dimensional imaging polarimetry. Proc. SPIE,2001,4377(8):175~185
22. J.Andrew, M.Nevis. Automated processing for streak tube imaging lidar data. Proc. SPIE,2003,5089(4):119~129
23.曹念文,刘文清,张玉钧.水下目标园偏振成像实验及最远成像距离的计算.中国激光,2000,27(2):150~154
24.曹念文,刘文清,张玉均.偏振成像技术提高成像清晰度、成像距离的定量研究.物理学报,2000,49(1):61~66
25. E.P.Zoge, A.Kokhannowsky. Analytucal solution to the optical transfer function of a scattening medium with large particles. Appl. Opt,1994,33(27):6547~6554
26. W.Liu, N.Cao, N.Takcuchi. Image enhancement by polarizing discrimination in turbid media. SPIE,1997,2081(12):3121~3125
27.李源慧,钟晓春,杨超.水下激光目标探测及其发展.光通信技术,2008,6(6):61~64
28.冯文波.蓝绿激光在水下传输中的应用.中南民族学院学报(自然科学版),2000,19(1):53~55
29. L.K.Smith. Investigation of the laser properties of Cr3+LiSr-CaF6. IEEE Journal of Quantum Electronics,1992,28(11):2612~2618
30.刘广荣,黄睿,金伟其,等.水下探测光电成像技术及其进展.光学技术,2004,30(6):732~737
31.王晓凯,李锋.改进的自适应中值滤波计算机工程与应用.2010,46(3)175~177
32.赵釜.基于Cyclone II系列FPGA的图像实时采集与预处理系统.重庆:重庆大学出版社,2009:57~68
33.刘雪明,张明德,孙小苗.同步扫描水下激光成像系统主要参数的分析.东南大学学报, 1999,1,(29):130~134
34.李田泽.激光的水下传输特性及其在测量中的应用.应用光学,1998,19(1):17~20
35.付伟.国外激光水下侦察设备发展综述.舰船电子对抗,2001,45(6):21~23
36.张昌达.激光侦查装备发展动向.激光与红外,1994,52(2):34~37
37.许大志,曹文熙,孙兆华.辽东湾海冰光衰减特性.海洋通报,2007,26(1):12~18
38. N.G.杰尔洛夫.海洋光学.北京:科学出版社,1981: 57~76
39.刘颂豪,李淳飞.光子学技术与应用.合肥:安徽科学技术出版社,2006:1933~1963
40.黄有为,金伟其,王霞,等.凝视型水下激光成像后向散射光理论模型研究.光学学报,2007,27(7):1192~1197
41. A .石丸.随机介质中波的传播和散射.北京:科学出版社, 1986:77~89
42. Y.Sasano, H.Shimizu, N.Takevchietal. Geometrical for Factor in the laser radar equation: an experimental determination. Appl. Opt,1979,18(23):3908~3910
43. T.E.Mcdonald, G.J.Yates, F.H.Cverna, et al. Rangegated imaging experiment using gated intensifiers. Proc. SPIE,1999,3642(47):142~148
44. M.C.Thomas, G.J.Yates. Image enhancement using a rang gated microchannel plate image intensifier(MCP II) video system with a 180ps FWHM shutter. Proc. SPIE,1995, 2551(26):174~180
45. G.Kervern.Imaging through scattering media Proc. SPIE,1992,1714(26):81~92
46. B.A.Swartz,J.D.Cummings. Laser range-gatedunder water imaging including polarizat ion discrim ination.Proc. SPIE,1991,1537(33):42~56
47.宋正方.应用大气光学基础.北京:气象出版社,1990:165~166
48.李丽,高稚允,王霞,等.水下距离选通成像系统后向散射光的计算.北京理工大学学报,2003,8,23(4)489~491
49.李自勤,李琦,王骐.由统计特性分析激光主动成像系统图像的噪声性质.中国激光,2004,31,(9):1082~1085
50. J.C.丹蒂.激光斑纹及有关现象.北京:科学出版社,1981:9~19
51. C.J.Oliver. Information from SAR images. J.Phys.D:ApplPhys,1991,24(9):1493~1514
52. J.S.Lee. Speckle suppression and analysis for synthetic aperture radar images. Optics Engineering,1986,25(5):636~643
53.韩宏伟,张晓晖.水下激光图像增强方法研究.激光与红外,2007,37(10):1106~1108
54.姚静,武文波,康停军.直方图均衡化的MATLAB实现.辽宁工程技术大学学报,2007,26(S2):61~62
55.张明谦,李雷.改进的中值滤波算法.兵工自动化,2007,26(5):77~80
56.曲延峰,徐健.有效去除图像中脉冲噪声的新型滤波算法.计算机辅助设计与图形学学报,2003,15(4):397~401
57.张燕.一种改进的快速中值滤波算法.安徽建筑工业学院学报(自然科学版),2008,16(4):21~35
58.李银华,路新惠.改进型自适应中值滤波算法在图像处理中的应用.郑州轻工业学院学报(自然科学版),2009,24(1):84~86
59.李自勤,王骐,李琦.激光成像雷达系统中散斑像的乘法模型及其滤除.中国激光,2003,30(8):719~720
2. L.R.Marshall. Rlue-green lasers plumb the mysteries of the deep. Laser Focus World,1993,4(4):185~197
3.孔捷,张保民.激光水下成像技术及其进展.光电子技术.2006,26(2):130~132
4.杨华勇,梁永辉.机载蓝绿激光水下目标探测技术的现状及前景.光机电信息,2003,12(1):6~10
5.朱仁慈.适合水下检查的激光成像.国外地质勘探技术,1993,4(2):14~23
6.李荣福,崔佳华,李桂娟.蓝绿脉冲激光用于海深及水下目标探测.声学与电子工程,1990,6(1):34~40
7.昌彦君,彭复员.水下激光成像的实验研究.实验室研究与探索.2009,28(3):19~20
8.李哲,邓甲昊,周卫平.水下激光探测技术及其进展.舰船电子工程.2008,28(12):8~11
9. A.GORDON. Use of laser scanning system on mobile underwater platforms. IEEE,1992,7803:202~205
10.刘雪明,谭志飞,胡正荣.水下成像系统及主要参数的理论计算.光电子激光,1999,9(1):71~72
11.刘雪明,张明德,孙小茜.水下激光成像系统的MTF分析.中国激光,1999,26(4):332~336
12. L.E.Mertens. Use of point spread and beam spread functions for analysis of imaging systems in water. J Opt. Soc. Am,1977,67(8):1105~1117
13.赵岩,翟百臣,王建立.激光距离选通成像同步控制系统的设计与实现.红外与激光工程.2005,34(5):527~529
14.孙健,周亚民.距离选通激光水下成像系统研究.舰船电子工程,2009,29(2):24~26
15.柏连发,张毅,陈钱.距离选通成像实现过程中若干问题的探讨.红外与激光工程,2009,38(1):58~61
16.昌彦君,彭复员.水下激光成像的实验研究.实验室研究与探索,2009,28(3):19~21
17.王祥科,张辉,郑志强.距离选通技术对后向散射抑制作用的理论分析.强激光与粒子束.2007,19(12):1998~2000
18.朱耘.激光水下三维成像技术及进展.应用光学,1999,20(4):22~26
19.李思宁,刘金波,李慧子.大视场情况下激光雷达距离方程的修正.中国激光,2006,33(suppl):108~110
20. D.Anthony, H.Gleckler, A.Gelbart, et al. Multispectral and hyperspectral 3D imaging lidar based upon the multiple slit streak tube imaging lidar. Proc. SPIE,2001,4377(8):328~335
21. D.Anthony, H.Gleckler, A.Gelbart. Three-dimensional imaging polarimetry. Proc. SPIE,2001,4377(8):175~185
22. J.Andrew, M.Nevis. Automated processing for streak tube imaging lidar data. Proc. SPIE,2003,5089(4):119~129
23.曹念文,刘文清,张玉钧.水下目标园偏振成像实验及最远成像距离的计算.中国激光,2000,27(2):150~154
24.曹念文,刘文清,张玉均.偏振成像技术提高成像清晰度、成像距离的定量研究.物理学报,2000,49(1):61~66
25. E.P.Zoge, A.Kokhannowsky. Analytucal solution to the optical transfer function of a scattening medium with large particles. Appl. Opt,1994,33(27):6547~6554
26. W.Liu, N.Cao, N.Takcuchi. Image enhancement by polarizing discrimination in turbid media. SPIE,1997,2081(12):3121~3125
27.李源慧,钟晓春,杨超.水下激光目标探测及其发展.光通信技术,2008,6(6):61~64
28.冯文波.蓝绿激光在水下传输中的应用.中南民族学院学报(自然科学版),2000,19(1):53~55
29. L.K.Smith. Investigation of the laser properties of Cr3+LiSr-CaF6. IEEE Journal of Quantum Electronics,1992,28(11):2612~2618
30.刘广荣,黄睿,金伟其,等.水下探测光电成像技术及其进展.光学技术,2004,30(6):732~737
31.王晓凯,李锋.改进的自适应中值滤波计算机工程与应用.2010,46(3)175~177
32.赵釜.基于Cyclone II系列FPGA的图像实时采集与预处理系统.重庆:重庆大学出版社,2009:57~68
33.刘雪明,张明德,孙小苗.同步扫描水下激光成像系统主要参数的分析.东南大学学报, 1999,1,(29):130~134
34.李田泽.激光的水下传输特性及其在测量中的应用.应用光学,1998,19(1):17~20
35.付伟.国外激光水下侦察设备发展综述.舰船电子对抗,2001,45(6):21~23
36.张昌达.激光侦查装备发展动向.激光与红外,1994,52(2):34~37
37.许大志,曹文熙,孙兆华.辽东湾海冰光衰减特性.海洋通报,2007,26(1):12~18
38. N.G.杰尔洛夫.海洋光学.北京:科学出版社,1981: 57~76
39.刘颂豪,李淳飞.光子学技术与应用.合肥:安徽科学技术出版社,2006:1933~1963
40.黄有为,金伟其,王霞,等.凝视型水下激光成像后向散射光理论模型研究.光学学报,2007,27(7):1192~1197
41. A .石丸.随机介质中波的传播和散射.北京:科学出版社, 1986:77~89
42. Y.Sasano, H.Shimizu, N.Takevchietal. Geometrical for Factor in the laser radar equation: an experimental determination. Appl. Opt,1979,18(23):3908~3910
43. T.E.Mcdonald, G.J.Yates, F.H.Cverna, et al. Rangegated imaging experiment using gated intensifiers. Proc. SPIE,1999,3642(47):142~148
44. M.C.Thomas, G.J.Yates. Image enhancement using a rang gated microchannel plate image intensifier(MCP II) video system with a 180ps FWHM shutter. Proc. SPIE,1995, 2551(26):174~180
45. G.Kervern.Imaging through scattering media Proc. SPIE,1992,1714(26):81~92
46. B.A.Swartz,J.D.Cummings. Laser range-gatedunder water imaging including polarizat ion discrim ination.Proc. SPIE,1991,1537(33):42~56
47.宋正方.应用大气光学基础.北京:气象出版社,1990:165~166
48.李丽,高稚允,王霞,等.水下距离选通成像系统后向散射光的计算.北京理工大学学报,2003,8,23(4)489~491
49.李自勤,李琦,王骐.由统计特性分析激光主动成像系统图像的噪声性质.中国激光,2004,31,(9):1082~1085
50. J.C.丹蒂.激光斑纹及有关现象.北京:科学出版社,1981:9~19
51. C.J.Oliver. Information from SAR images. J.Phys.D:ApplPhys,1991,24(9):1493~1514
52. J.S.Lee. Speckle suppression and analysis for synthetic aperture radar images. Optics Engineering,1986,25(5):636~643
53.韩宏伟,张晓晖.水下激光图像增强方法研究.激光与红外,2007,37(10):1106~1108
54.姚静,武文波,康停军.直方图均衡化的MATLAB实现.辽宁工程技术大学学报,2007,26(S2):61~62
55.张明谦,李雷.改进的中值滤波算法.兵工自动化,2007,26(5):77~80
56.曲延峰,徐健.有效去除图像中脉冲噪声的新型滤波算法.计算机辅助设计与图形学学报,2003,15(4):397~401
57.张燕.一种改进的快速中值滤波算法.安徽建筑工业学院学报(自然科学版),2008,16(4):21~35
58.李银华,路新惠.改进型自适应中值滤波算法在图像处理中的应用.郑州轻工业学院学报(自然科学版),2009,24(1):84~86
59.李自勤,王骐,李琦.激光成像雷达系统中散斑像的乘法模型及其滤除.中国激光,2003,30(8):719~720