摘 要：图像修复的目的在于根据已知的图像信息训练提取特征，再依据提取的有效特征恢复图像受损部分的像素。随着深度学习技术的发展，数字图像修复技术可以帮助人们解决更加复杂的问题。文章梳理了该领域近年来的研究现状：首先概述了传统图像修复方法以及存在的缺陷，然后对比分析修复效果有显著提高的图像修复方法，最后总结当前数字图像修复技术存在的缺陷并对将来的研究方向进行展望。
关键词：图像修复；计算机视觉；自编码网络
DOI:10.19850/j.cnki.2096-4706.2022.04.010
中图法分类号：TP391.4 文献标识码：A 文章编号：2096-4706（2022）04-0038-03
Research Progress of Digital Image Inpainting Methods
WANG Ke
(College of Physics and Information Engineering, Fuzhou University, Fuzhou 350108, China)
Abstract: The purpose of image inpainting is to train and extract features according to known image information, and then restore the damaged part pixels of the image based on the extracted effective features. With the development of deep learning technology, digital image inpainting technology can help people solve more complex problems. This paper combs the current state of research in this field in recent years: firstly, it outlines the traditional image inpainting methods and their defects, then it compares and analyzes the image inpainting methods that have significantly improved the inpainting effect, finally, it summarizes the defects of current digital image inpainting techniques and provides an outlook on future research directions.
Keywords: image inpainting; computer vision; auto encoder network
参考文献：
[1] BERTALMIO M，SAPIRO G，CASELLES V，et al. Image inpainting [EB/OL].[2021.12.03].https://dl.acm.org/doi/pdf/10.1145/344779.344972.
[2] CHAN T F，SHEN J H. Nontexture Inpainting by CurvatureDriven Diffusions [J].Journal of visual communication and image representation，2001，12（4）：436-449.
[3] SHEN J J，CHAN T F. Mathematical models for local nontexture inpaintings [J].SIAM Journal on Applied Mathematics，2002，62（3）：1019-1043.
[4] CRIMINISI A，PEREZ P，TOYAMA K. Region filling and object removal by exemplar-based image inpainting [J].IEEE Transactions on image processing，2004，13（9）：1200-1212.
[5] BARNES C，SHECHTMAN E，FINKELSTEIN A，et al. PatchMatch： A randomized correspondence algorithm for structural image editing [J].ACM Transactions on Graphic，2009，28（3）：24.
[6] RUMELHART D E，HINTON G E，WILLIAMS R J.Learning internal representations by error propagation [J].Readings in Cognitive Science，1988：399-421.
[7] PATHAK D，KRAHENBUHL P，DONAHUE J，et al. Context encoders： Feature learning by inpainting [C]//2016 IEEE Conference on Computer Vision and Pattern Recognition （CVPR）.Las Vegas：IEEE，2016：2536-2544.
[8] GOODFELLOW I J，POUGET-ABADIE J，MIRZA M，et al. Generative adversarial nets [J].Proceedings of the 27th International Conference on Neural Information Processing Systems，2014，2：2627-2628.
[9] ZHAO L，MO Q H，LIN S H，et al. UCTGAN： Diverse Image Inpainting Based on Unsupervised Cross-Space Translation [C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition （CVPR）.Seattle：IEEE，2020：5740-5749.
[10] LIU H Y，WAN Z Y，HUANG W，et al. PD-GAN：Probabilistic Diverse GAN for Image Inpainting [J/OL].arXiv：2105.02201 [cs.CV].[2021-12-14].https://arxiv.org/abs/2105.02201v1.
[11] YI Z L，TANG Q，AZIZI S，et al. Contextual Residual Aggregation for Ultra High-Resolution Image Inpainting [C]//2020 IEEE/CVF Conference on Computer Vision and Pattern Recognition （CVPR）.Seattle：IEEE，2020：7505-7514.
作者简介：王柯（1994—），女，汉族，河南周口人，硕士研究生在读，研究方向：图像修复与图像生成。