-
摘要: 针对液晶可调滤波片高光谱成像系统记录动态场景的成像特点，提出一种图-谱结合的压缩感知高光谱视频图像复原方法。首先，通过前景目标检测获得运动前景目标的高光谱图像，实现运动前景目标与背景区域分离，并根据前景目标检测结果将背景区域划分为运动区域（被前景目标遮挡区域）与静止区域（未被前景目标遮挡区域）。然后，基于高光谱图像空间维、光谱维相关性，对静止区域进行字典学习获得稀疏先验信息，结合压缩感知理论用于运动区域恢复，得到完整的背景区域高光谱图像。最后，将运动前景目标高光谱图像与背景区域高光谱图像相结合，得到高光谱视频图像。实验结果表明：本文提出的高光谱视频图像复原方法在峰值信噪比和视觉效果上都要优于现有算法，峰值信噪比平均提高5 dB以上。Abstract: In this paper, a new graph-spectral hyperspectral video restoration method regarding the imaging characteristics of dynamic scenes recorded by liquid crystal tunable filter hyperspectral imaging system is proposed. Firstly, the hyperspectral image of the moving foreground target is obtained by the foreground target detection, and the moving foreground target is separated from the background region. Then the background region is divided into the motion region which is obscured by the foreground target and the still region which is not obscured by the foreground target according to the foreground target detection result. Based on the correlation of the spatial dimension and spectral dimension of the hyperspectral image, dictionary learning is performed on the still region to obtain sparse prior information. Combined with compressed sensing theory for motion region recovery, a complete background region hyperspectral image is obtained. Finally, the moving foreground target hyperspectral image is combined with the background region hyperspectral image to obtain a hyperspectral video image. The experimental results show that the proposed method of hyperspectral video image restoration outperforms the existing algorithm in terms of peak signal-to-noise ratio and visual effect, and the peak signal-to-noise ratio is increased by an average of more than 5 dB.
-
图 2 基于液晶可调滤波片高光谱成像系统采集的图像示意图
Figure 2. An illustration of images captured by hyperspectral imaging system based on liquid crystal tunable filter
图 3 An illustration of images captured by hyperspectral imaging system based on liquid crystal tunable filter
Figure 3. Flowchart of hyperspectral video restoration
图 6 高光谱视频的部分复原结果对比图
Figure 6. Comparison of some results of hyperspectral video restoration
表 1 本文算法与文献[7]中方法的PSNR值对比
Table 1. Output PSNR comparison of the proposed method and the method in Ref.[7]
(Unit:dB) 高光谱视频帧数 本文方法 文献[7]中的方法 第1帧 31.44 25.76 第2帧 31.53 26.30 第3帧 31.90 27.38 第4帧 32.63 27.78 第5帧 32.88 28.15 第6帧 32.71 28.05 第7帧 32.54 27.64 第8帧 32.80 26.76 第9帧 32.35 25.92 第10帧 31.61 23.72 平均 32.24 26.75 
下载: 导出CSV
-
[1] 马晨光, 曹汛, 戴琼海, 等.高分辨率光谱视频采集研究[J].电子学报, 2015, 43(4):783-790. doi: 10.3969/j.issn.0372-2112.2015.04.022MA CH G, CAO X, DAI Q H, et al.. Research on high resolution hyperspectral capture technique[J]. Chinese Journal of Electronics, 2015, 43(4):783-790.(in Chinese) doi: 10.3969/j.issn.0372-2112.2015.04.022 [2] 付立婷, 邓河, 刘春红.新型高光谱图像快速实时目标检测与分类方法[J].光学学报, 2017, 37(2):314-322. http://www.cnki.com.cn/Article/CJFDTOTAL-GXXB201702038.htmFU L T, DENG H, LIU CH H. Novel fast real-time target detection and classification algorithms for hyperspectral imagery[J]. Acta Optica Sinica, 2017, 37(2):314-322.(in Chinese) http://www.cnki.com.cn/Article/CJFDTOTAL-GXXB201702038.htm [3] 梅风华, 李超, 张玉鑫.光谱成像技术在海域目标探测中的应用[J].中国光学, 2017, 10(6):708-718. http://www.chineseoptics.net.cn/CN/abstract/abstract9565.shtmlMEI F H, LI CH, ZH Y X. Application of spectral imaging technology in maritime target detection[J]. Chinese Optics, 2017, 10(6):708-718.(in Chinese) http://www.chineseoptics.net.cn/CN/abstract/abstract9565.shtml [4] CLAUDIA V C, DIANA F G, HENRY A F. Sparse representations of dynamic scenes for compressive spectral video sensing[J]. Dyna Revista De La Facultad De Minas, 2016, 83(195):42-51. [5] WAGADARIKAR A, PITSIANIS N, SUN X, et al.. Video rate spectral imaging using a coded aperture snapshot spectral imager[J]. Optics Express, 2009, 17(8):6368-6388. doi: 10.1364/OE.17.006368 [6] CAO X, DU H, DAI Q H, et al.. A prism-mask system for multispectral video acquisition[J]. IEEE Transactions on Pattern Analysis & Machine Intelligence, 2011, 33(12):2423-2435. http://dl.acm.org/citation.cfm?id=2068459.2068565 [7] AJMAL M, RICHARD H. Hyperspectral video restoration using optical flow and sparse coding[J]. Optics Express, 2012, 20(10):10658-10673. doi: 10.1364/OE.20.010658 [8] BARNICH O, VANDROOGENBROECK M. ViBe:a universal background subtraction algorithm for video sequences[J]. IEEE Transactions on Image Processing, 2011, 20(6):1709-1724. doi: 10.1109/TIP.2010.2101613 [9] 丁洁, 况立群, 韩燮.基于ViBe的运动目标检测改进方法[J].计算机工程与设计, 2017, 38(2):374-378. http://d.old.wanfangdata.com.cn/Periodical/jsjgcysj201702018DING J, KUANG L Q, HAN X. Improved moving object detection method based on ViBe[J]. Computer Engineering and Design, 2017, 38(2):374-378.(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/jsjgcysj201702018 [10] 胡昭华, 张维新, 王珏, 等.基于改进ViBe的运动目标检测算法[J].电子技术应用, 2017, 43(4):129-132. http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=dzjsyy201704034HU ZH H, ZHANG W X, WANG J, et al.. Moving object detection algorithm based on improved ViBe[J]. Application of Electronic Technique, 2017, 43(4):129-132.(in Chinese) http://www.wanfangdata.com.cn/details/detail.do?_type=perio&id=dzjsyy201704034 [11] 戴琼海, 付长军, 季向阳.压缩感知研究[J].计算机学报, 2011, 34(3):3425-3434. http://d.old.wanfangdata.com.cn/Periodical/dianzixb200905028DAI Q H, FU CH J, JI X Y. Research on compressed sensing[J]. Chinese Journal of Computers, 2011, 34(3):3425-3434.(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/dianzixb200905028 [12] 朱明, 高文, 郭立强.压缩感知理论在图像处理领域的应用[J].中国光学, 2011, 4(5):441-447. doi: 10.3969/j.issn.2095-1531.2011.05.006ZHU M, GAO W, GUO L Q. Application of compressed sensing theory in image processing[J]. Chinese Optics, 2011, 4(5):441-447.(in Chinese) doi: 10.3969/j.issn.2095-1531.2011.05.006 [13] 王忠良, 冯燕, 肖华, 等.高光谱图像的分布式压缩感知成像与重构[J].光学精密工程, 2015, 23(4):1131-1137. http://d.old.wanfangdata.com.cn/Periodical/gxjmgc201504028WANG ZH L, FENG Y, XIAO H, et al.. Distributed compressive sensing imaging and reconstruction of hyperspectral imagery[J]. Opt. Precision Eng., 2015, 23(4):1131-1137.(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/gxjmgc201504028 [14] 杨鹰, 孔玲君, 刘真.基于压缩感知的多光谱图像去马赛克算法[J].液晶与显示, 2017, 32(1):56-61. http://d.old.wanfangdata.com.cn/Periodical/yjyxs201701010YANG Y, KONG L J, LIU ZH. Multi-spectral demosaicking method based on compressive sensing[J]. Chinese Journal of Liquid Crystals and Displays, 2017, 32(1):56-61.(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/yjyxs201701010 [15] 汪琪, 马灵玲, 唐伶俐, 等.基于光谱稀疏模型的高光谱压缩感知重构[J].红外与毫米波学报, 2016, 35(6):723-730. http://d.old.wanfangdata.com.cn/Periodical/hwyhmb201606015WANG Q, MA L L, TANG L L, et al.. Hyperspectral compressive sensing reconstruction based on spectral sparse model[J]. Journal of Infrared and Millimeter Waves, 2016, 35(6):723-730.(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/hwyhmb201606015 [16] 王晗, 王阿川, 苍圣.基于压缩感知的高光谱遥感影像重构方法研究[J].液晶与显示, 2017, 32(3):219-226. http://d.old.wanfangdata.com.cn/Periodical/yjyxs201703009WANG H, WANG A CH, C SH. Hyperspectral remote sensing image reconstruction method based on compressive sensing[J]. Chinese Journal of Liquid Crystals and Displays, 2017, 32(3):219-226.(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/yjyxs201703009 [17] GAN L. Block compressed sensing of natural images[C]. Proceedings of the 15th International Conference on Digital Signal Processing, Cardiff, Wales, 2007: 403-406. [18] DONG W SH, ZHANG L, SHI G M, et al.. Image deblurring and super-resolution by adaptive sparse domain selection and adaptive regularization[J]. IEEE Transactions on Image Processing, 2011, 20(7):1838-1857. doi: 10.1109/TIP.2011.2108306 [19] 首照宇, 吴广祥, 张彤.基于PCA子字典学习的图像超分辨率重建[J].计算机工程与设计, 2015, 36(11):3025-3029. http://d.old.wanfangdata.com.cn/Periodical/jsjgcysj201511029SHOU ZH Y, WU G X, ZHANG T. Image super-resolution reconstruction via PCA sub-dictionaries learning[J]. Computer Engineering and Design, 2015, 36(11):3025-3029.(in Chinese) http://d.old.wanfangdata.com.cn/Periodical/jsjgcysj201511029 [20] WRIGHT S J, NOWAK R D, FIGUEIREDO M A T. Sparse reconstruction by separable approximation[J]. IEEE Transactions on Signal Processing, 2009, 57(7):2479-2493. doi: 10.1109/TSP.2009.2016892 -
下载:
点击查看大图
图(6) / 表(1)
计量
- 文章访问数: 2055
- HTML全文浏览量: 631
- PDF下载量: 168
- 被引次数: 0
