留言板

尊敬的读者、作者、审稿人, 关于本刊的投稿、审稿、编辑和出版的任何问题, 您可以本页添加留言。我们将尽快给您答复。谢谢您的支持!

姓名
邮箱
手机号码
标题
留言内容
验证码

基于改进引导滤波器的多光谱去马赛克方法

齐海超 宋延嵩 张博 梁宗林 闫纲琦 薛佳音 张轶群 任斌

齐海超, 宋延嵩, 张博, 梁宗林, 闫纲琦, 薛佳音, 张轶群, 任斌. 基于改进引导滤波器的多光谱去马赛克方法[J]. 中国光学(中英文), 2023, 16(5): 1056-1065. doi: 10.37188/CO.2022-0231
引用本文: 齐海超, 宋延嵩, 张博, 梁宗林, 闫纲琦, 薛佳音, 张轶群, 任斌. 基于改进引导滤波器的多光谱去马赛克方法[J]. 中国光学(中英文), 2023, 16(5): 1056-1065. doi: 10.37188/CO.2022-0231
QI Hai-chao, SONG Yan-song, ZHANG Bo, LIANG Zong-lin, YAN Gang-qi, XUE Jia-yin, ZHANG Yi-qun, REN Bin. Multispectral demosaicing method based on an improved guided filter[J]. Chinese Optics, 2023, 16(5): 1056-1065. doi: 10.37188/CO.2022-0231
Citation: QI Hai-chao, SONG Yan-song, ZHANG Bo, LIANG Zong-lin, YAN Gang-qi, XUE Jia-yin, ZHANG Yi-qun, REN Bin. Multispectral demosaicing method based on an improved guided filter[J]. Chinese Optics, 2023, 16(5): 1056-1065. doi: 10.37188/CO.2022-0231

基于改进引导滤波器的多光谱去马赛克方法

基金项目: 国家重点研发计划资助项目(No. 2022YFB3902500);国家自然科学基金资助项目(No. U2141231);吉林省自然科学基金(No. 202002036JC);鹏城实验室重大攻关项目(No. PCL2021A03-1)
详细信息
    作者简介:

    齐海超(1997—),男,吉林长春人,硕士研究生,2019年于长春理工大学获得学士学位,主要从事图像处理方面的研究。E-mail:qihaichao2019@163.com

    宋延嵩(1983—),男,吉林长春人,博士,研究员,博士生导师,2006年、2009年、2014年于长春理工大学分别获得学士、硕士及博士学位,主要研究方向为空间激光通信技术。E-mail: songyansong2006@126.com

  • 中图分类号: TP391.9

Multispectral demosaicing method based on an improved guided filter

Funds: Supported by National Key R & D Program of China (No. 2022YFB3902500); National Natural Science Foundation of China (No.U2141231); the Natural Science Foundation of Jilin Province (No. 202002036JC); The Major Key Project of PCL (No. PCL2021A03-1)
More Information
  • 摘要:

    为了更好地保留多光谱去马赛克图像中的高频信息,本文提出了一种基于改进引导滤波器的多光谱图像去马赛克方法。首先,基于自回归模型对相邻像素点间的强相关性进行建模,在每个像素处渐进估计其模型参数,通过最小化局部窗口内的估计误差,得到最优估计值来插值采样密集波段G,并生成高质量的引导图像;然后,引入加窗固有变分系数到惩罚因子中,得到具有边缘感知能力的加权引导滤波器并重建其余稀疏采样波段。最后,使用CAVE数据集和TokyoTech数据集进行仿真。实验结果表明:相较于主流的5波段多光谱图像去马赛克方法,本方法重建图像的峰值信噪比和结构相似度在CAVE数据集和TokyoTech数据集上分别提高了3.40%,2.02%,1.34%,0.30%和6.11%,5.95%,2.28%,1.42%,且更好地保留了原始图像的局部结构和颜色信息,减少了边缘伪影和噪声现象的出现。

     

  • 图 1  5波段MSFA(a)二叉树分裂过程及(b)排列模式

    Figure 1.  (a) Binary tree splitting process and (b) arrangement of Five-band MSFA

    图 2  邻域T内像素排列

    Figure 2.  Pixel arrangement in neighborhood T

    图 3  水平-垂直方向上的自回归模型

    Figure 3.  Autoregressive model in the horizontal-vertical direction

    图 4  模型参数估计

    Figure 4.  Estimation of model parameter

    图 5  基于加权引导滤波的去马赛克流程

    Figure 5.  Demosaicing process based on weight-guided filtering

    图 6  不同算法Balloons场景重建图像对比

    Figure 6.  Comparison of Balloons images reconstructed by different algorithms

    图 7  不同算法CD场景重建图像对比

    Figure 7.  Comparison of CD images reconstructed by different algorithms

    图 8  不同算法Party场景重建图像对比

    Figure 8.  Comparison of party images reconstructed by different algorithms

    表  1  CAVE数据集上3种方法的客观评价指标

    Table  1.   Objective evaluation metrics of the three methods on the CAVE dataset

    CAVEsRGB PSNR/dBsRGB SSIMCIEDE 2000
    GFAPMIDProGFAPMIDProGFAPMIDPro
    Balloons41.6242.6843.110.98590.99160.99361.181.060.99
    Clay37.3337.6338.690.87580.88170.88521.070.940.89
    Beers41.5742.1843.520.98160.98680.98941.251.281.07
    Lemons42.9142.8742.910.97490.98050.98221.111.081.03
    Peppers42.1442.0842.520.97150.98010.98050.890.780.73
    Feathers35.6435.9435.990.94130.95930.96142.302.102.04
    Flowers38.9341.1842.500.95230.97780.98241.260.910.83
    Paints36.1634.8836.320.96960.97290.97972.402.432.17
    Apples45.2445.1045.680.98380.98750.98860.770.750.70
    Toys38.8341.2942.770.96660.98450.98911.240.900.80
    Avg40.0440.5841.400.96030.97030.97321.351.221.13
    下载: 导出CSV

    表  2  TokyoTech数据集上3种方法的客观评价指标

    Table  2.   Objective evaluation metrics of the three methods on the TokyoTech dataset

    TokyoTechsRGB PSNR/dBsRGB SSIMCIEDE 2000
    GFAPMIDProGFAPMIDProGFAPMIDPro
    Butterfly37.5338.9540.440.95960.96780.98101.601.421.17
    Butterfly338.4242.9441.990.94870.97770.97931.370.910.84
    Butterfly440.6340.7342.370.96910.95900.98271.121.230.87
    CD32.2032.7832.870.94500.95800.96291.971.721.65
    Character37.7437.4538.140.96730.97360.98351.831.941.73
    Cloth34.1835.0035.870.93210.94810.95733.343.222.75
    Color39.2238.6241.360.97820.96300.98951.742.001.47
    Colorchart42.8344.7947.800.98190.98470.99410.920.770.55
    Fan232.6833.2934.090.92570.94260.96292.632.342.04
    Party32.7933.4535.780.93660.95090.96932.061.661.36
    Avg36.8237.8039.070.95440.96250.97621.861.721.44
    下载: 导出CSV

    表  3  不同方法在两种数据集上的运行时间

    Table  3.   Running times of different methods on the two datasets (s)

    数据集GFAPMIDPro
    CAVE1.490.711.33
    TokyoTech1.360.561.29
    下载: 导出CSV
  • [1] ORTEGA S, HALICEK M, FABELO H, et al. Hyperspectral and multispectral imaging in digital and computational pathology: a systematic review[J]. Biomedical Optics Express, 2020, 11(6): 3195-3233. doi: 10.1364/BOE.386338
    [2] 王成, 刘宾, 周楚, 等. 窄带LED照明的多光谱显微成像系统研究[J]. 中国激光,2020,47(12):1207006. doi: 10.3788/CJL202047.1207006

    WANG CH, LIU B, ZHOU CH, et al. Multispectral microimaging system with narrowband LED illumination[J]. Chinese Journal of Lasers, 2020, 47(12): 1207006. (in Chinese) doi: 10.3788/CJL202047.1207006
    [3] 唐凌宇, 葛明锋, 董文飞. 全自动推扫式高光谱显微成像系统设计与研究[J]. 中国光学,2021,14(6):1486-1494. doi: 10.37188/CO.2021-0040

    TANG L Y, GE M F, DONG W F. Design and research of fully automatic push-broom hyperspectral microscopic imaging system[J]. Chinese Optics, 2021, 14(6): 1486-1494. (in Chinese) doi: 10.37188/CO.2021-0040
    [4] SU W H, SUN D W. Multispectral imaging for plant food quality analysis and visualization[J]. Comprehensive Reviews in Food Science and Food Safety, 2018, 17(1): 220-239. doi: 10.1111/1541-4337.12317
    [5] CHAMBINO L L, SILVA J S, BERNARDINO A. Multispectral facial recognition: a review[J]. IEEE Access, 2020, 8: 207871-207883. doi: 10.1109/ACCESS.2020.3037451
    [6] WU F, JING X Y, FENG Y J, et al. Spectrum-aware discriminative deep feature learning for multi-spectral face recognition[J]. Pattern Recognition, 2021, 111: 107632. doi: 10.1016/j.patcog.2020.107632
    [7] 李云辉. 压缩光谱成像系统中物理实现架构研究综述[J]. 中国光学(中英文),2022,15(5):929-945. doi: 10.37188/CO.2022-0104

    LI Y H. Review of physical implementation architecture in compressive spectral imaging system[J]. Chinese Optics, 2022, 15(5): 929-945. (in Chinese) doi: 10.37188/CO.2022-0104
    [8] 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. doi: 10.3788/YJYXS20173201.0056

    YANG 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) doi: 10.3788/YJYXS20173201.0056
    [9] HABTEGEBRIAL T A, REIS G, STRICKER D. Deep convolutional networks for snapshot hypercpectral demosaicking[C]. 2019 10th Workshop on Hyperspectral Imaging and Signal Processing: Evolution in Remote Sensing (WHISPERS), IEEE, 2019: 1-5.
    [10] FENG K, ZHAO Y Q, CHAN J C W, et al. Mosaic convolution-attention network for demosaicing multispectral filter array images[J]. IEEE Transactions on Computational Imaging, 2021, 7: 864-878. doi: 10.1109/TCI.2021.3102052
    [11] 肖树林, 胡长虹, 高路尧, 等. 像元映射变分辨率光谱成像重构[J]. 中国光学(中英文),2022,15(5):1045-1054.

    XIAO SH L, HU CH H, GAO L Y, et al. Pixel mapping variable-resolution spectral imaging reconstruction[J]. Chinese Optics, 2022, 15(5): 1045-1054. (in Chinese)
    [12] MIAO L D, QI H R. The design and evaluation of a generic method for generating mosaicked multispectral filter arrays[J]. IEEE Transactions on Image Processing, 2006, 15(9): 2780-2791. doi: 10.1109/TIP.2006.877315
    [13] MIAO L D, QI H R, RAMANATH R, et al. Binary tree-based generic demosaicking algorithm for multispectral filter arrays[J]. IEEE Transactions on Image Processing, 2006, 15(11): 3550-3558. doi: 10.1109/TIP.2006.877476
    [14] GUPTA M, RAM M. Weighted bilinear interpolation based generic multispectral image demosaicking method[J]. Journal of Graphic Era University, 2019, 7(2): 108-118.
    [15] GUPTA M, RATHI V, GOYAL P. Adaptive and progressive multispectral image demosaicking[J]. IEEE Transactions on Computational Imaging, 2022, 8: 69-80. doi: 10.1109/TCI.2022.3140554
    [16] 孙帮勇, 袁年曾, 胡炳樑. 一种八谱段滤光片成像系统设计[J]. 光子学报,2020,49(5):0511001. doi: 10.3788/gzxb20204905.0511001

    SUN B Y, YUAN N Z, HU B L. Design of an eight-band filter imaging system[J]. Acta Photonica Sinica, 2020, 49(5): 0511001. (in Chinese) doi: 10.3788/gzxb20204905.0511001
    [17] RATHI V, GOYAL P. Generic multispectral Demosaicking based on directional interpolation[J]. IEEE Access, 2022, 10: 64715-64728. doi: 10.1109/ACCESS.2022.3182493
    [18] MONNO Y, TANAKA M, OKUTOMI M. Multispectral demosaicking using guided filter[J]. Proceedings of SPIE, 2012, 8299: 82990O. doi: 10.1117/12.906168
    [19] 任杰, 刘家瑛, 白蔚, 等. 基于隐式分段自回归模型的图像插值算法[J]. 软件学报,2012,23(5):1248-1259. doi: 10.3724/SP.J.1001.2012.04049

    REN J, LIU J Y, BAI W, et al. Image interpolation algorithm based on implicit piecewise autoregressive model[J]. Journal of Software, 2012, 23(5): 1248-1259. (in Chinese) doi: 10.3724/SP.J.1001.2012.04049
    [20] ZHANG X J, WU X L. Image interpolation by adaptive 2-D autoregressive modeling and soft-decision estimation[J]. IEEE Transactions on Image Processing, 2008, 17(6): 887-896. doi: 10.1109/TIP.2008.924279
    [21] HE K M, SUN J, TANG X O. Guided image filtering[J]. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2013, 35(6): 1397-1409. doi: 10.1109/TPAMI.2012.213
    [22] LI ZH G, ZHENG J H, ZHU Z J, et al. Weighted guided image filtering[J]. IEEE Transactions on Image Processing, 2015, 24(1): 120-129. doi: 10.1109/TIP.2014.2371234
    [23] XU L, YAN Q, XIA Y, et al. Structure extraction from texture via relative total variation[J]. ACM Transactions on Graphics, 2012, 31(6): 139.
    [24] 路陆, 姜鑫, 杨锦程, 等. 基于自适应引导滤波的红外图像细节增强[J]. 液晶与显示,2022,37(9):1182-1189. doi: 10.37188/CJLCD.2022-0024

    LU L, JIANG X, YANG J CH, et al. Adaptive guided filtering based infrared image detail enhancement[J]. Chinese Journal of Liquid Crystals and Displays, 2022, 37(9): 1182-1189. (in Chinese) doi: 10.37188/CJLCD.2022-0024
    [25] YASUMA F, MITSUNAGA T, ISO D, et al. Generalized assorted pixel camera: Postcapture control of resolution, dynamic range, and spectrum[J]. IEEE Transactions on Image Processing, 2010, 19(9): 2241-2253. doi: 10.1109/TIP.2010.2046811
    [26] MONNO Y, KIKUCHI S, TANAKA M, et al. A practical one-shot multispectral imaging system using a single image sensor[J]. IEEE Transactions on Image Processing, 2015, 24(10): 3048-3059. doi: 10.1109/TIP.2015.2436342
    [27] PARK J I, LEE M H, GROSSBERG M D, et al. . Multispectral imaging using multiplexed illumination[C]. 2007 IEEE 11th International Conference on Computer Vision, IEEE, 2007: 1-8.
    [28] SARA U, AKTER M, UDDIN M S. Image quality assessment through FSIM, SSIM, MSE and PSNR—a comparative study[J]. Journal of Computer and Communications, 2019, 7(3): 8-18. doi: 10.4236/jcc.2019.73002
    [29] SHARMA G, WU W CH, DALAL E N. The CIEDE2000 color-difference formula: Implementation notes, supplementary test data, and mathematical observations[J]. Color Research &Application, 2005, 30(1): 21-30.
    [30] 贾停停, 王慧琴, 王可, 等. 相位相关性增强的自适应低重叠率多光谱图像快速拼接算法[J]. 液晶与显示,2022,37(4):483-493. doi: 10.37188/CJLCD.2021-0294

    JIA T T, WANG H Q, WANG K, et al. Adaptive low overlap multispectral image fast mosaic algorithm based on phase correlation enhancement[J]. Chinese Journal of Liquid Crystals and Displays, 2022, 37(4): 483-493. (in Chinese) doi: 10.37188/CJLCD.2021-0294
  • 加载中
图(8) / 表(3)
计量
  • 文章访问数:  474
  • HTML全文浏览量:  315
  • PDF下载量:  224
  • 被引次数: 0
出版历程
  • 收稿日期:  2022-11-13
  • 修回日期:  2022-12-12
  • 网络出版日期:  2023-04-17

目录

    /

    返回文章
    返回