[1]郑心草,孙忠贵.图像超分中双三次插值的非局部拓展[J].南京师范大学学报(工程技术版),2023,23(03):053-59.[doi:10.3969/j.issn.1672-1292.2023.03.007]
　Zheng Xincao,Sun Zhonggui.Non-local Extension of Bicubic Interpolation in Image Super-resolution[J].Journal of Nanjing Normal University(Engineering and Technology),2023,23(03):053-59.[doi:10.3969/j.issn.1672-1292.2023.03.007]
点击复制

图像超分中双三次插值的非局部拓展
分享到：

南京师范大学学报（工程技术版）[ISSN:1006-6977/CN:61-1281/TN]

卷:
23卷
期数:
2023年03期
页码:
053-59
栏目:
计算机科学与技术
出版日期:
2023-09-15

文章信息/Info

Title:
Non-local Extension of Bicubic Interpolation in Image Super-resolution
文章编号:
1672-1292(2023)03-0053-07
作者:
郑心草1; 2; 孙忠贵2
(1.聊城大学季羡林学院,山东 聊城 252000)
(2.聊城大学数学科学学院,山东 聊城 252000)
Author(s):
Zheng Xincao1; 2; Sun Zhonggui2
(1.Ji Xianlin Honors School, Liaocheng University, Liaocheng 252000, China)
(2.School of Mathematical Sciences, Liaocheng University, Liaocheng 252000, China)
关键词:
图像超分; 双三次插值; 非局部均值; 周期性
Keywords:
image super-resolution;  bicubic;  non-local means;  periodicity
分类号:
TP391.41
DOI:
10.3969/j.issn.1672-1292.2023.03.007
文献标志码:
A
摘要:
因具备强大的细节刻画能力,双三次插值已成为图像超分中的常用算法. 由于其借助在空间距离上与当前像素距离最近的 16 个像素构造插值基函数,故双三次插值本质上属于一种局部算法. 这也意味着该算法的插值过程尚不能有效利用图像周期性(非局部性),从而致使其细节保持能力仍存在进一步提升空间. 针对这一问题,通过对原插值基函数施加非局部权重修正,在一个更大范围内选取更多像素对当前像素的灰度值进行估计,实现了经典双三次插值算法的非局部拓展. 在灰度图像和彩色图像两个不同场景上进行超分实验,主观视觉效果和客观量化指标均表明所提算法的有效性.
Abstract:
Benefiting from its powerful ability in detail preseriving, bicubic interpolation has become a common algorithm in image super-resolution. The base function for bicubic interpolation is constructed by the 16 pixels that are closest to the current pixel in space, so the bicubic interpolation belongs to a local algorithm. This means that the interpolation process in this algorithm is not able to take advantage of the periodicity(non-locality)in natural images, and thus its performance still has room to be further improved. Based on this considerations, this paper introduces non-local weight correction into the definition of the base function to select more candidate pixels in a wider area to estimate the gray value of the current pixel. As a result, a non-local version of the typical bicubic interpolation is achieved. The experimental results suggest that it can reconstruct higher quality results both quantitatively and perceptually.

参考文献/References:

[1]ZHANG K B,GAO X B,TAO D C,et al. Single image super-resolution with non-local means and steering kernel regression[J]. IEEE Transactions on Image Processing,2012,21(11):4544-4556.
[2]杜婉君,孙忠贵. 跨模态空域自适应联合均值滤波器[J]. 南京师范大学学报(工程技术版),2022,22(1):52-58.
[3]ZHAO F,HU Z F. Image super resolution reconstruction algorithm based on convolution neural network[J]. Journal of Physics:Conference Series,2021,2025:012033.
[4]LI M,NGUYEN T Q. Markov random field model-based edge-directed image interpolation[J]. IEEE Transactions on Image Processing,2008,17(7):1121-1128.
[5]CHENG Y,WANG Y,LIU Q. Research on digital image scaling based on bicubic filter algorithm[C]//Proceedings of the 2018 IEEE 3rd International Conference on Image,Vision and Computing(ICIVC). Chongqing,China:IEEE,2018.
[6]GAO X B,WANG Q,LI X L,et al. Zernike-moment-based image super resolution[J]. IEEE Transactions on Image Processing,2011,20(10):2738-2747.
[7]NAZREN A R A,YAAKOB S N,NGADIRAN R,et al. Improving iterative back projection super resolution model via anisotropic diffusion edge enhancement[C]//Proceedings of the 2016 International Conference on Robotics,Automation and Sciences(ICORAS). Melaka,Malaysia:IEEE,2016.
[8]ZHANG Y L,WEI D L,QIN C,et al. Context reasoning attention network for image super-resolution[C]//Proceedings of the 2021 IEEE/CVF International Conference on Computer Vision(ICCV). Montreal,Canada:IEEE,2021.
[9]TONG T,LI G,LIU X J,et al. Image super-resolution using dense skip connections[C]//Proceedings of the 2017 IEEE International Conference on Computer Vision(ICCV). Venice,Italy:IEEE,2017.
[10]张宇苏,吴小俊,李辉,等. 基于无监督深度学习的红外图像与可见光图像融合算法[J]. 南京师范大学学报(工程技术版),2023,23(1):1-9.
[11]JIANG J J,CHEN C,MA J Y,et al. SRLSP:a face image super-resolution algorithm using smooth regression with local structure prior[J]. IEEE Transactions on Multimedia,2017,19(1):27-40.
[12]BUADES A,COLL B,MOREL J M. A non-local algorithm for image denoising[C]//Proceedings of the 2005 IEEE Computer Society Conference on Computer Vision and Pattern Recognition(CVPR'05). San Diego,USA:IEEE,2005.
[13]KUMAR B K S. Image denoising based on non-local means filter and its method noise thresholding[J]. Signal Image and Video Processing,2013,7(6):1211-1227.
[14]SUN Z G,LYU M Q,LI J,et al. A local-nonlocal mathematical morphology[J]. Neurocomputing,2022,495:51-61.
[15]李兵,刘全升,徐家伟,等. 去除混合噪音的一种新方法[J]. 中国科学:信息科学,2010,40(9):1165-1175.
[16]WU W W,LIU F L,ZHANG Y B,et al. Non-local low-rank cube-based tensor factorization for spectral CT reconstruction[J]. IEEE Transactions on Medical Imaging,2019,38(4):1079-1093.
[17]KEYS R. Cubic convolution interpolation for digital image processing[J]. IEEE Transactions on Acoustics,Speech and Signal Processing,1981,29(6):1153-1160.
[18]裴银祥. 快速非局部均值图像去噪算法[D]. 郑州:郑州大学,2014.
[19]WANG X T,SHEN S S,SHI G M,et al. Iterative non-local means filter for salt and pepper noise removal[J]. Journal of Visual Communication and Image Representation,2016,38:440-450.
[20]周涛,刘珊,董雅丽,等. 肺部肿瘤跨模态图像融合的并行分解自适应融合模型[J]. 中国图象图形学报,2023,28(1):221-233.
[21]吴岸聪,林城梽,郑伟诗. 面向跨模态行人重识别的单模态自监督信息挖掘[J]. 中国图象图形学报,2022,27(10):2843-2859.

备注/Memo

备注/Memo:
收稿日期:2023-03-09.
基金项目:国家自然科学基金项目(11801249)、山东省自然科学基金项目(ZR2020MF040)、聊城大学开放课题项目(319462207-1).
通讯作者:孙忠贵,博士,教授,研究方向:图像处理、机器学习. E-mail:altlp@163.com
更新日期/Last Update: 2023-09-15