一种减少空间调制快拍成像测偏仪伪信息的方法
|
摘要
空间调制快拍成像测偏技术能将偏振信息编码到一幅干涉图中,通过一次测量同时获取目标全部斯托克斯参量.针对空间调制快拍成像测偏技术中,目标像的高频部分对偏振信息通道产生串扰,导致重构的偏振信息包含有伪信息和频域重构目标像的空间分辨率低等问题,提出一种消除伪信息和获取全分辨率目标像的方法.该方法通过对两次快拍测量获得的干涉图进行简单加减运算,便可获得探测目标清晰的纯图像和高信噪比的偏振分量干涉图.本文对该方法进行了详细的理论分析,并通过计算机仿真和搭建实验平台,验证了该方法的可行性.这为空间调制快拍成像测偏技术获取全分辨率目标像和高质量重构偏振信息提供了新思路.
Spatially-modulated snapshot imaging polarimeter can encode four Stokes parameters(S_0, S_1, S_2 and S_3) into a single interferogram and allow the instantaneous measurement of polarization from a single snapshot. However, the reconstructed polarization information contains aliasing signal, and the reconstructed intensity images suffer low spatial resolution because of the crosstalk between high frequency components of the image and frequency domain filtering for the polarization channels. In this paper, we propose an image superposition and subtraction method to mitigate the aliasing problem and to recover the image resolution. The two interferograms acquired from two snapshot measurements are superposed to obtain the intensity image(S_0 component) of an object without the polarization components because the phases of the polarization components in the two interferograms are opposite. In comparison with the intensity of each of the original interferograms, the intensity of S_0 component increases twice and its spatial resolution improves up to a maximum value offered by the instrument. Then a subtraction between the two interferograms is performed to derive the pure interference fringes while the intensity image vanishes. The intensity of the pure interference fringes also increases twice compared with that of each original interferogram because phases of the interference terms in original interferograms are opposite. The polarization images(S_1, S_2 and S_3 components) can be reconstructed from the pure interference fringes, and do not include crosstalk signals between the high frequency components of the intensity image.The theoretical basis of the method is presented through a detailed analysis. Its feasibility is verified by both computer simulation and experiment. The simulation results show that the otherness and the structural similarity index between the input and reconstructed intensity images is zero and 1, respectively, indicating a perfect reconstruction of S_0. The results also make it clear that the pure interference fringes do not include any component of intensity image, and thus the reconstructed polarization information does not contain any crosstalk signals. Moreover, the experimental results are in accordance with the theoretical expectation and the computer simulations. This research provides a novel means for spatially-modulated snapshot imaging polarization technology to obtain full-resolution object images and high-quality reconstructed polarization information.
Spatially-modulated snapshot imaging polarimeter can encode four Stokes parameters(S_0, S_1, S_2 and S_3) into a single interferogram and allow the instantaneous measurement of polarization from a single snapshot. However, the reconstructed polarization information contains aliasing signal, and the reconstructed intensity images suffer low spatial resolution because of the crosstalk between high frequency components of the image and frequency domain filtering for the polarization channels. In this paper, we propose an image superposition and subtraction method to mitigate the aliasing problem and to recover the image resolution. The two interferograms acquired from two snapshot measurements are superposed to obtain the intensity image(S_0 component) of an object without the polarization components because the phases of the polarization components in the two interferograms are opposite. In comparison with the intensity of each of the original interferograms, the intensity of S_0 component increases twice and its spatial resolution improves up to a maximum value offered by the instrument. Then a subtraction between the two interferograms is performed to derive the pure interference fringes while the intensity image vanishes. The intensity of the pure interference fringes also increases twice compared with that of each original interferogram because phases of the interference terms in original interferograms are opposite. The polarization images(S_1, S_2 and S_3 components) can be reconstructed from the pure interference fringes, and do not include crosstalk signals between the high frequency components of the intensity image.The theoretical basis of the method is presented through a detailed analysis. Its feasibility is verified by both computer simulation and experiment. The simulation results show that the otherness and the structural similarity index between the input and reconstructed intensity images is zero and 1, respectively, indicating a perfect reconstruction of S_0. The results also make it clear that the pure interference fringes do not include any component of intensity image, and thus the reconstructed polarization information does not contain any crosstalk signals. Moreover, the experimental results are in accordance with the theoretical expectation and the computer simulations. This research provides a novel means for spatially-modulated snapshot imaging polarization technology to obtain full-resolution object images and high-quality reconstructed polarization information.
引文
[1]Snik F,Craven-Jones J,Escuti M,Fineschid S,Harringtone D,Martinof A D,Mawetg D,Riedih J,Tyo J S 2014 Proc.SPIE 9099,Polarization:Measurement,Analysis,and Remote Sensing XI Baltimore,Maryland,United States,March 24–28,2014 p90990B
[2]Forward S,Gribble A,Alali S,Andras A L,Vitkin I A2017 Sci.Reports 7 11958
[3]Tyo J S,Goldstein D L,Chenault D B,Shaw J A 2006Appl.Opt.45 5453
[4]Alali S,Vitkin A 2015 J.Biomed.Opt.20 0611041
[5].Oka K,Kaneko T 2003 Opt.Express 11 1510
[6]Luo H T 2008 Ph.D.Dissertation(Tucson:University of Arizona)
[7]Cao Q Z 2014 Ph.D.Dissertation(Xi’an:xi’an Jiaotong University)(in Chinese)[曹奇志2014博士学位论文(西安:西安交通大学)]
[8]Cao Q Z,Zhang C M,De Hoog E 2012 Appl.Opt.515791
[9]Jian X H,Zhang C M,Zhao B C 2007 Acta Phys.Sin.56 824(in Chinese)[简小华,张淳民,赵葆常2007物理学报56 824]
[10]Peng Z H,Zhang C M,Zhao B C,Li Y C,Wu F Q 2006Acta Phys.Sin.55 6374(in Chinese)[彭志红,张淳民,赵葆常,李英才,吴福全2006物理学报55 6374]
[11]Qu Y,Zhang C M,Wang D Y,Tian P B,Bai W G,Zhang X Y,Zhang P,Dai H S,Wu Q M 2013 Int.J.Remote Sens.34 3938
[12]Yuan Z L,Zhang C M,Zhao B C 2007 Acta Phys.Sin.56 6413(in Chinese)[彭志红,张淳民,赵葆常2007物理学报56 6413]
[13]Oka K,Haga Y,Komaki Y 2013 Proceedings Volume8873,Polarization Science and Remote Sensing VI San Diego,California,United States September 27,2013p88730R-1
[14]Oka K,Haga Y,Michida H 2015 Proceedings Volume9613,Polarization Science and Remote Sensing VII San Diego,California,United States,September 1,2015p96130E-9
[15]Oka,K,Kodai S,Hiroshi M 2017 Proceedings Volume10407,Polarization Science and Remote Sensing VIII San Diego,California,United States,September 7,2017p104070M
[16]Hu Q Y,Yang W F,Hu Y D,Hong J 2015 Acta Opt.Sin.2 144(in Chinese)[胡巧云,杨伟锋,胡亚东,洪津2015光学学报2 144]
[17]Cao Q Z,Zhang J,De Hoog E,Lu Y,Hu B Q,Li W G,Li J Y,Fan D X,Deng T,Yan Y 2016 Acta Phys.Sin.65 050702(in Chinese)[曹奇志,张晶,Edward De Hoog,卢远,胡宝清,李武钢,李建映,樊东鑫,邓婷,阎妍2016物理学报65 050702]
[18]Cao Q Z,Zhang C M,Zhang J,Kang Y Q 2014 Optik125 3380
[19]Cao Q Z,Zhang J,De Hoog E,Zhang C M 2016 Appl.Opt.55 954
[20]Kudenov M,Escuti M,Dereniak E,Oka K 2011 Appl.Opt.50 2283
[21]Kudenov M W 2009 Ph.D.Dissertation(Tucson:The University of Arizona)
[22]HoréA,Ziou D 2010 Proceedings of IEEE Conference on Pattern Recognition Istanbul,Turkey October 7,2010p2366
[23]https://www.mathworks.com/help/images/ref/ssim.html?search Highlight=SSIM&s_tid=doc_srchtitle[2017-11-8]
[24]Wang Z,Bovik C,Sheikh R,Simoncelli P 2004 IEEE Trans.on Image Process.13 600
[2]Forward S,Gribble A,Alali S,Andras A L,Vitkin I A2017 Sci.Reports 7 11958
[3]Tyo J S,Goldstein D L,Chenault D B,Shaw J A 2006Appl.Opt.45 5453
[4]Alali S,Vitkin A 2015 J.Biomed.Opt.20 0611041
[5].Oka K,Kaneko T 2003 Opt.Express 11 1510
[6]Luo H T 2008 Ph.D.Dissertation(Tucson:University of Arizona)
[7]Cao Q Z 2014 Ph.D.Dissertation(Xi’an:xi’an Jiaotong University)(in Chinese)[曹奇志2014博士学位论文(西安:西安交通大学)]
[8]Cao Q Z,Zhang C M,De Hoog E 2012 Appl.Opt.515791
[9]Jian X H,Zhang C M,Zhao B C 2007 Acta Phys.Sin.56 824(in Chinese)[简小华,张淳民,赵葆常2007物理学报56 824]
[10]Peng Z H,Zhang C M,Zhao B C,Li Y C,Wu F Q 2006Acta Phys.Sin.55 6374(in Chinese)[彭志红,张淳民,赵葆常,李英才,吴福全2006物理学报55 6374]
[11]Qu Y,Zhang C M,Wang D Y,Tian P B,Bai W G,Zhang X Y,Zhang P,Dai H S,Wu Q M 2013 Int.J.Remote Sens.34 3938
[12]Yuan Z L,Zhang C M,Zhao B C 2007 Acta Phys.Sin.56 6413(in Chinese)[彭志红,张淳民,赵葆常2007物理学报56 6413]
[13]Oka K,Haga Y,Komaki Y 2013 Proceedings Volume8873,Polarization Science and Remote Sensing VI San Diego,California,United States September 27,2013p88730R-1
[14]Oka K,Haga Y,Michida H 2015 Proceedings Volume9613,Polarization Science and Remote Sensing VII San Diego,California,United States,September 1,2015p96130E-9
[15]Oka,K,Kodai S,Hiroshi M 2017 Proceedings Volume10407,Polarization Science and Remote Sensing VIII San Diego,California,United States,September 7,2017p104070M
[16]Hu Q Y,Yang W F,Hu Y D,Hong J 2015 Acta Opt.Sin.2 144(in Chinese)[胡巧云,杨伟锋,胡亚东,洪津2015光学学报2 144]
[17]Cao Q Z,Zhang J,De Hoog E,Lu Y,Hu B Q,Li W G,Li J Y,Fan D X,Deng T,Yan Y 2016 Acta Phys.Sin.65 050702(in Chinese)[曹奇志,张晶,Edward De Hoog,卢远,胡宝清,李武钢,李建映,樊东鑫,邓婷,阎妍2016物理学报65 050702]
[18]Cao Q Z,Zhang C M,Zhang J,Kang Y Q 2014 Optik125 3380
[19]Cao Q Z,Zhang J,De Hoog E,Zhang C M 2016 Appl.Opt.55 954
[20]Kudenov M,Escuti M,Dereniak E,Oka K 2011 Appl.Opt.50 2283
[21]Kudenov M W 2009 Ph.D.Dissertation(Tucson:The University of Arizona)
[22]HoréA,Ziou D 2010 Proceedings of IEEE Conference on Pattern Recognition Istanbul,Turkey October 7,2010p2366
[23]https://www.mathworks.com/help/images/ref/ssim.html?search Highlight=SSIM&s_tid=doc_srchtitle[2017-11-8]
[24]Wang Z,Bovik C,Sheikh R,Simoncelli P 2004 IEEE Trans.on Image Process.13 600
© 2004-2018 中国地质图书馆版权所有 京ICP备05064691号 京公网安备11010802017129号 地址：北京市海淀区学院路29号 邮编：100083 电话：办公室：(+86 10)66554848；文献借阅、咨询服务、科技查新：66554700 |