Image blind restoration based on edge constraint and norm ratio

Abstract

Abstract: The image blur caused by motion has always been a challenging problem. The image prior information used in the classical moving image blind restoration algorithm is often too simple. The sparse theory has a good restoration effect, and it usually estimates the point spread function directly using the gradient edges of the degraded image. But the gradient edge of the degraded image contains many weak edges and pseudo edges, which can affect the estimation of the point spread function. To solve the above problems, a blind image restoration algorithm based on edge constraint and norm ratio is proposed. First, the edge of the degraded image is constrained to obtain strong edge structure of the image, which improves the accuracy of point spread function estimation. Then, the sparse penalty constraint of norm ratio is constructed for the clear image to be estimated, and the obtained strong edge information is combined with the constructed norm ratio penalty constraint to guide the restoration of point spread function. When restoring the point spread function, the coarse to fine multiscale iterative estimation of the point spread function makes the iterative maximum scale more accurate. Finally, the image is solved by non-blind deconvolution to restore it. The algorithm combines the edge information of the degraded image with the constructed norm ratio penalty constraint to guide the restoration of the point spread function. It can suppress a large number of artifacts generated in the process of image restoration. Experimental results show that the algorithm has a good effect on restoring the edge details of the image, which can get a high-quality restored image.

Key words: image blind restoration, multiscale, norm ratio, strong edge, point spread function

CLC Number:

TP391

ZHAO Xiao-qiang, WANG Tao. Image blind restoration based on edge constraint and norm ratio[J]. Journal of Lanzhou University of Technology, 2024, 50(1): 76-83.

References

[1] 李晖晖,鱼轮,张良,等.暗通道约束和交替方向乘子法优化的湍流图像盲复原 [J].西北工业大学学报,2018,36(1):103-109.
[2] 刘旭.图像去模糊算法研究 [D].杭州:浙江大学,2021.
[3] 黄正源,谢维成,黄化入,等.图像运动模糊还原技术综述 [J].计算机工程与应用,2020,56(24):28-34.
[4] 赵宏伟,霍东升,王洁,等.基于显著性检测的害虫图像分类 [J].吉林大学学报(工学版),2021,51(6):2174-2181.
[5] 陈刚琦.基于卷积神经网络的高原鼠兔图像检测与分割方法研究 [D].兰州:兰州理工大学,2020.
[6] LIU S S,ZHANG H,MAO Z H,et al.Target detection method based on HRM feature extracting and SVM [J].Foreign Electronic Measurement Technology,2014,33(10):38-41.
[7] 张云,李岚.基于级联卷积神经网络的人脸特征点识别算法实现 [J].兰州理工大学学报,2020,46(3):105-109.
[8] AYERS G R,DAINTY J C.Iterative blind deconvolution method and its applications [J].Optics Letters,1988,13(7):547-549.
[9] GUO Z H,KAN Z H,SHAO F Q,et al.Anisotropic regularization method for electrical capacitance tomography image reconstruction [J].Chinese Journal of Scientific Instrument,2013,34(102):2854-2859.
[10] FAN W R,WANG H X,HAO K H.Two-step iterative TV regularization algorithm for image reconstruction of electrical impedance tomography [J].Chinese Journal of Scientific Instrument,2012,33(3):615-630.
[11] CHAN T F,WONG C K.Total variation blind deconvolution [J].IEEE Transactions on Image Processing,1998,7(3):370-375.
[12] FERGUS R,SINGH B,HERTZMANN A,et al.Removing camera shake from a single photograph [J].ACM Transactions on Graphics,2006,25(25):787-794.
[13] LEVIN A,WEISS Y,DURAND F,et al.Understanding and evaluating blind deconvolution algorithms [C]//IEEE Conference on Computer Vision and Pattern Recognition.Piscataway:IEEE,2009:1964-1971.
[14] LEVIN A,WEISS Y,DURAND F,et al.Efficient marginal likehood optimization in blind deconvolution [C]//Proceedings of 2011 IEEE Computer Society Conference on Computer Vision and Pattern Recognition Workshops.Providence:RI IEEE Computer Society,2011:2657-2664.
[15] KRISHNAN D,FERGUS R.Fast image deconvolution using hyper-Laplacian priors [C]//International Conference on Neural Information Processing Systems.Vancouver:[s.n.],2009:1033-1041.
[16] CAI J F,JI H,LIU C,et al.Blind motion deblurring from a single image using sparse approximation [C]//IEEE Conference on Computer Vision and Pattern Recognition.Miami,FL:IEEE,2009:104-111.
[17] HU J B,YANG M H.Single image deblurring with adaptive dictionary learning [C]//17th IEEEInternational Conference on Image Processing.Hong Kong:IEEE,2010:1169-1172.
[18] JIA C,EVANS B L.Patch-based image deconvolution via joint modeling of sparse priors [C]//18th IEEE International Conference on Information Processing.Brussels:IEEE,2011:681-684.
[19] KRISHNAN D,TAY T,FERGUS R.Blind deconvolution using a normalized sparsity measure [C]//Computer Vision and Pattern Recognition.Providence,RI:[s.n.],2011:233-240.
[20] BECK A,TEBOULLE M.A fast iterative shrinkage-thresholding algorithm for linear inverse problems [J].SIAM Journal on Imaging Sciences,2009,2(1):183-202.
[21] LEVIN A,FERGUS R,DURAND F,et al.Image and depth from a conventional camera with a coded aperture [J].ACM Transactions on Graphics,2007,26(3):701-709.
[22] XU L,YAN Q,XIA Y,et al.Structure extraction from texture via relative total variation [J].Acm Transactionson Graphics,2012,21(6):439-445.
[23] KOTERA J,SROUBEK F,MILANFAR P.Blind deconvolution using alternating maximum a posteriori estimation with heavy-tailed priors [C]//International Conference on Computer Analysis of Images and Patterns.Berlin:Springer,2013:59-66.