搜索

x

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

贝叶斯迭代联合双边滤波的散焦图像快速盲复原

尹诗白 王卫星 王一斌 李大鹏 邓箴

引用本文:
Citation:

贝叶斯迭代联合双边滤波的散焦图像快速盲复原

尹诗白, 王卫星, 王一斌, 李大鹏, 邓箴

Fast Bayesian blind restoration for single defocus image with iterative joint bilateral filters

Yin Shi-Bai, Wang Wei-Xing, Wang Yi-Bin, Li Da-Peng, Deng Zhen
PDF
导出引用
  • 实现有效的单幅散焦图像盲复原对军事及地质勘测领域的清晰图像获取具有极为重要的意义.常用算法存在计算量大、振铃及噪声敏感的问题,为此本文提出了贝叶斯框架下迭代双边滤波器的快速盲复原算法.它首先用基于深度信息的盲去卷积结果估计点扩散函数的概率模型,进而通过贝叶斯理论构建合理的盲复原最小优化问题;然后推理分析最小优化问题的求解实质,得出双边滤波器快速求解最小优化问题的结论;最后设计迭代联合双边滤波器的求解方式,即利用一次双边滤波器求解的复原结果设计联合双边滤波器的指导图,再将其作为优化问题的输入,迭代实施求解.实验结果表明:该算法能有效抑制振铃,减少计算量,去除噪声,85%图像的像素误差平均值低于0.03,较常用盲去卷积法在同一误差区间的复原成功率提高了19%,运行时间缩短了约78%,能有效用于单幅散焦图像盲复原的实际工程实践中.
    It is significant to realize effective defocus image restoration for acquiring clear image in military and geological examination field. Most of existing algorithms have the problems of large computational cost, ringing and noise sensitivity, hence a novel approach by iterative joint bilateral filtering under Bayesian framework is proposed. Firstly, it utilizes defocus image depth estimation to compute the point spread function in the Bayesian framework. Then a minimum optimization problem is built to represent the blind restoration problem. After inferencing the solution procedure of the minimum optimization problem, we find that the joint bilateral filters can be used to search the optimal solution, which not only simplifies the searching procedure but also reduces the computational cost. Finally, an iterative joint bilateral filtering is designed to realize the image restoration. That means that the original restored image obtained from the bilateral filtering is used to design the guide image for the joint bilateral filters, and the guide image will serve as the input of the optimization problem for acquiring the better optimal result. This procedure is repeated until convergence. The experimental results indicate that this method can yield the ringing, reduce the computational cost, and remove the noise. Generally speaking, the average pixel error of 85% images is under 0.03, which has improved 19% comparing with the same error rang of existing algorithms, and 78% shorter than those of compared algorithms. It can be used in the engineering practice of blind restoration for single defocus image.
      通信作者: 尹诗白, shibaiyin@swufe.edu.cn
    • 基金项目: 国家自然科学基金重大项目(批准号:91218301)、国家自然科学基金青年科学基金(批准号:61502396)、中央高校基本科研业务费(批准号:JBK150503,JBK160135)和宁夏自然科学基金(批准号:NZ15054)资助的课题.
      Corresponding author: Yin Shi-Bai, shibaiyin@swufe.edu.cn
    • Funds: Project supported by the Major Program of the National Natural Science Foundation of China (Grant No. 91218301), the Young Scientists Fund of the National Natural Science Foundation of China (Grant No. 61502396), the Fundamental Research Fund for the Central Universities, China (Grant Nos. JBK150503, JBK160135), and the Natural Science Foundation of Ningxia, China (Grant No. NZ15054).
    [1]

    Schuon S, Diepold K 2009 Acta Astronaut. 64 1050

    [2]

    Gupta P, Mehra R 2015 Int. J. Comput. Appl. 130 20

    [3]

    Escande P, Weiss P, Malgouyres F 2013 J. Phys. 2013 012004

    [4]

    Galdran A, Pardo D, Picón A, Alvarez-Gila A 2015 J. Visual Commun. Image Represent. 26 132

    [5]

    Jin Z L, Han J, Zhang Y, Bai L F 2014 Acta Phys. Sin. 63 069501 (in Chinese)[金左轮, 韩静, 张毅, 柏连发2014物理学报63 069501]

    [6]

    Shi M Z, Xu T F, Liang J, Li X M 2013 Acta Phys. Sin. 62 174204 (in Chinese)[石明珠, 许廷发, 梁炯, 李相民2013物理学报62 174204]

    [7]

    Li X N, Huang H Y, Jia X N, Ma S L 2015 Acta Phys. Sin. 64 134102 (in Chinese)[李鑫楠, 黄贺艳, 贾小宁, 马驷良2015物理学报64 134102]

    [8]

    Lu H M, Xu M, Li X 2014 Acta Opt. Sin. 2014 081002 (in Chinese)[卢惠民, 徐明, 李迅2014光学学报2014 081002]

    [9]

    Tai Y W, Brown M S 2009 In Proc of 16th IEEE International Conf on Image Processing Cairo, Egypt, November 7-10, 2009 p1797

    [10]

    Cai J F, Ji H, Liu C, Shen Z 2009 J. Comput. Phys. 228 5057

    [11]

    Kundur D, Hatzinakos D 1996 IEEE Trans. Signal Process. 13 43

    [12]

    Fahmy M F, Raheem G M A, Mohamed U S, Fahmy O F 2012 J. Signal Inf. Process. 3 98

    [13]

    Almeida M S, Figueiredo M A 2013 IEEE Trans. Image Process. 22 2751

    [14]

    Schmidt U, Schelten K, Roth S 2011 In Proc of 16th IEEE Conf on Computer Vision and Pattern Recognition Colorado, United State, June 21-23, 2011 p2625

    [15]

    Likas A C, Galatsanos N P 2004 IEEE Trans. Signal Process. 52 2222

    [16]

    Zhang H, Wipf D, Zhang Y 2014 IEEE Trans. Pattern Anal. Mach. Intell. 36 1628

    [17]

    Song C, Deng H, Gao H, Zhang H, Zuo W 2016 Neurocomputing 197 95

    [18]

    Cao Y, Fang S, Wang F 2011 In Proc. of 6th International Conf. on Image and Graphics Beijing, China, October 24-26, 2011 p168

    [19]

    Elad M 2005 International Conf on Scale-Space Theories in Computer Vision in Scale Space and PDE Methods in Computer Vision (Berlin Heidelberg:Springer) pp217-229

    [20]

    Huynh T Q, Ghanbari M 2008 Electron. Lett. 44 800

    [21]

    Levin A, Weiss Y, Durand F, Freeman W T 2011 In Proc of 18th IEEE Conf on Computer Vision and Pattern Recognition Colorado, United State, June 21-23, 2011 p2657

  • [1]

    Schuon S, Diepold K 2009 Acta Astronaut. 64 1050

    [2]

    Gupta P, Mehra R 2015 Int. J. Comput. Appl. 130 20

    [3]

    Escande P, Weiss P, Malgouyres F 2013 J. Phys. 2013 012004

    [4]

    Galdran A, Pardo D, Picón A, Alvarez-Gila A 2015 J. Visual Commun. Image Represent. 26 132

    [5]

    Jin Z L, Han J, Zhang Y, Bai L F 2014 Acta Phys. Sin. 63 069501 (in Chinese)[金左轮, 韩静, 张毅, 柏连发2014物理学报63 069501]

    [6]

    Shi M Z, Xu T F, Liang J, Li X M 2013 Acta Phys. Sin. 62 174204 (in Chinese)[石明珠, 许廷发, 梁炯, 李相民2013物理学报62 174204]

    [7]

    Li X N, Huang H Y, Jia X N, Ma S L 2015 Acta Phys. Sin. 64 134102 (in Chinese)[李鑫楠, 黄贺艳, 贾小宁, 马驷良2015物理学报64 134102]

    [8]

    Lu H M, Xu M, Li X 2014 Acta Opt. Sin. 2014 081002 (in Chinese)[卢惠民, 徐明, 李迅2014光学学报2014 081002]

    [9]

    Tai Y W, Brown M S 2009 In Proc of 16th IEEE International Conf on Image Processing Cairo, Egypt, November 7-10, 2009 p1797

    [10]

    Cai J F, Ji H, Liu C, Shen Z 2009 J. Comput. Phys. 228 5057

    [11]

    Kundur D, Hatzinakos D 1996 IEEE Trans. Signal Process. 13 43

    [12]

    Fahmy M F, Raheem G M A, Mohamed U S, Fahmy O F 2012 J. Signal Inf. Process. 3 98

    [13]

    Almeida M S, Figueiredo M A 2013 IEEE Trans. Image Process. 22 2751

    [14]

    Schmidt U, Schelten K, Roth S 2011 In Proc of 16th IEEE Conf on Computer Vision and Pattern Recognition Colorado, United State, June 21-23, 2011 p2625

    [15]

    Likas A C, Galatsanos N P 2004 IEEE Trans. Signal Process. 52 2222

    [16]

    Zhang H, Wipf D, Zhang Y 2014 IEEE Trans. Pattern Anal. Mach. Intell. 36 1628

    [17]

    Song C, Deng H, Gao H, Zhang H, Zuo W 2016 Neurocomputing 197 95

    [18]

    Cao Y, Fang S, Wang F 2011 In Proc. of 6th International Conf. on Image and Graphics Beijing, China, October 24-26, 2011 p168

    [19]

    Elad M 2005 International Conf on Scale-Space Theories in Computer Vision in Scale Space and PDE Methods in Computer Vision (Berlin Heidelberg:Springer) pp217-229

    [20]

    Huynh T Q, Ghanbari M 2008 Electron. Lett. 44 800

    [21]

    Levin A, Weiss Y, Durand F, Freeman W T 2011 In Proc of 18th IEEE Conf on Computer Vision and Pattern Recognition Colorado, United State, June 21-23, 2011 p2657

  • [1] 伍静, 崔春凤, 欧阳滔, 唐超. 基于贝叶斯算法的5-7环缺陷石墨烯纳米带热电性能优化设计. 物理学报, 2023, 72(4): 047201. doi: 10.7498/aps.72.20222135
    [2] 郝望, 段睿, 杨坤德. 联合简正波水波和底波频散特性的贝叶斯地声参数反演. 物理学报, 2023, 72(5): 054303. doi: 10.7498/aps.72.20221717
    [3] 娄月申, 郭文军. 贝叶斯深度神经网络对于核质量预测的研究. 物理学报, 2022, 71(10): 102101. doi: 10.7498/aps.71.20212387
    [4] 杨棣, 王元美, 李军刚. 贝叶斯频率估计中频率的先验分布对有色噪声作用的影响. 物理学报, 2018, 67(6): 060301. doi: 10.7498/aps.67.20171911
    [5] 姚军财, 刘贵忠. 基于图像内容视觉感知的图像质量客观评价方法. 物理学报, 2018, 67(10): 108702. doi: 10.7498/aps.67.20180168
    [6] 李倩倩, 阳凡林, 张凯, 郑炳祥. 不确定海洋环境中基于贝叶斯理论的声源运动参数估计方法. 物理学报, 2016, 65(16): 164304. doi: 10.7498/aps.65.164304
    [7] 郭苗苗, 王昱婧, 徐桂芝, Griffin Milsap, Nitish V. Thakor, Nathan Crone. 时变动态贝叶斯网络模型及其在皮层脑电网络连接中的应用. 物理学报, 2016, 65(3): 038702. doi: 10.7498/aps.65.038702
    [8] 李慧, 赵琳, 李亮. 基于贝叶斯压缩感知的周跳探测与修复方法. 物理学报, 2016, 65(24): 249101. doi: 10.7498/aps.65.249101
    [9] 文方青, 张弓, 贲德. 基于块稀疏贝叶斯学习的多任务压缩感知重构算法. 物理学报, 2015, 64(7): 070201. doi: 10.7498/aps.64.070201
    [10] 李鑫楠, 黄贺艳, 贾小宁, 马驷良. 基于指导滤波的图像盲复原算法. 物理学报, 2015, 64(13): 134202. doi: 10.7498/aps.64.134202
    [11] 王保宪, 赵保军, 唐林波, 王水根, 吴京辉. 基于双向稀疏表示的鲁棒目标跟踪算法. 物理学报, 2014, 63(23): 234201. doi: 10.7498/aps.63.234201
    [12] 赵佳, 喻莉, 李静茹. 社交网络中基于贝叶斯和半环代数模型的节点影响力计算机理. 物理学报, 2013, 62(13): 130201. doi: 10.7498/aps.62.130201
    [13] 陈明惠, 丁志华, 王成, 宋成利. 基于法布里-珀罗调谐滤波器的傅里叶域锁模扫频激光光源. 物理学报, 2013, 62(6): 068703. doi: 10.7498/aps.62.068703
    [14] 王娇, 周云辉, 黄玉清, 江虹. 基于贝叶斯网络的认知引擎设计与重配置. 物理学报, 2013, 62(3): 038402. doi: 10.7498/aps.62.038402
    [15] 来志, 郭亮, 李小珍, 党文佳. 机器视觉在激光干涉测量d31中的应用. 物理学报, 2013, 62(18): 184207. doi: 10.7498/aps.62.184207
    [16] 李斐. 相位差图像复原技术研究. 物理学报, 2012, 61(23): 230203. doi: 10.7498/aps.61.230203
    [17] 李金才, 黄斌, 彭宇行. 一种改进的用于合成孔径雷达图像相干斑抑制的双边滤波参数配置方法. 物理学报, 2012, 61(18): 189501. doi: 10.7498/aps.61.189501
    [18] 颜鹏程, 侯威, 钱忠华, 何文平, 孙建安. 基于贝叶斯理论的全球海温异常对500 hPa 温度场的影响分析. 物理学报, 2012, 61(13): 139202. doi: 10.7498/aps.61.139202
    [19] 郝崇清, 王江, 邓斌, 魏熙乐. 基于稀疏贝叶斯学习的复杂网络拓扑估计. 物理学报, 2012, 61(14): 148901. doi: 10.7498/aps.61.148901
    [20] 杨 光, 陈桂英, 祁胜文, 郝召锋, 田建国, 张春平. 非均匀输入图像对基于细菌视紫红质膜的新事物滤波器输出特性的数值模拟. 物理学报, 2007, 56(12): 6954-6960. doi: 10.7498/aps.56.6954
计量
  • 文章访问数:  5287
  • PDF下载量:  296
  • 被引次数: 0
出版历程
  • 收稿日期:  2016-07-09
  • 修回日期:  2016-09-11
  • 刊出日期:  2016-12-05

/

返回文章
返回