[1]杨元英,王安志*,何淋艳,等.基于生成式对抗网络的图像修复研究进展[J].计算机技术与发展,2022,32(02):75-81.[doi:10. 3969 / j. issn. 1673-629X. 2022. 02. 012]
　YANG Yuan-ying,WANG An-zhi*,HE Lin-yan,et al.Advances in Image Inpainting Based on Generative Adversarial Networks[J].,2022,32(02):75-81.[doi:10. 3969 / j. issn. 1673-629X. 2022. 02. 012]
点击复制

基于生成式对抗网络的图像修复研究进展()
分享到：

《计算机技术与发展》[ISSN:1006-6977/CN:61-1281/TN]

卷:
32
期数:
2022年02期
页码:
75-81
栏目:
图形与图像
出版日期:
2022-02-10

文章信息/Info

Title:
Advances in Image Inpainting Based on Generative Adversarial Networks
文章编号:
1673-629X(2022)02-0075-07
作者:
杨元英; 王安志*; 何淋艳; 任春洪; 欧卫华
贵州师范大学 大数据与计算机科学学院,贵州 贵阳 550025
Author(s):
YANG Yuan-ying; WANG An-zhi* ; HE Lin-yan; REN Chun-hong; OU Wei-hua
School of Big Data and Computer Science,Guizhou Normal University,Guiyang 550025,China
关键词:
生成式对抗网络; 图像修复; 生成器; 判别器; 自编码器
Keywords:
generative adversarial network; image inpainting; generator; discriminator; autoencoder
分类号:
TP391. 41
DOI:
10. 3969 / j. issn. 1673-629X. 2022. 02. 012
摘要:
图像修复是图像处理的一个重要问题,目的是利用计算机视觉技术自动恢复退化图像中损坏或丢失的部分,被广泛应用于影视特技制作、图像编辑、数字化文物保护等领域。 近几年,以生成式对抗网络( GAN) 为代表的深度学习技术在计算机视觉和图像处理领域大获成功,基于 GAN 的图像修复逐渐成为主流,受到了广泛关注。 针对图像修复的关键问题,文章对 GAN 和基于 GAN 的修复方法进行理论分析,首先整理分析了传统的基于人工特征的经典图像修复方法,其次总结了近年来基于 GAN 的代表性图像修复算法,并进行归纳分类,探讨了各类方法的特点和局限性。 然后对图像修复模型常用的评价指标和公开数据集进行整理和分析,最后阐述了图像修复面临的挑战,对图像修复技术未来的发展方向进行展望。
Abstract:
Inpainting is an important problem in image processing,which aims to use computer vision technology to automatically restoredamaged or lost parts in degraded images. It is widely used in film and television special effects production,image editing,digital heritageprotection. In recent years,the deep learning technology represented by generative adversarial network ( GAN) has achieved great successin the field of computer vision and image processing. And GANs based inpainting methods have gradually been widely concerned.Therefore,to cope with the key issues of image inpainting,we make a theoretical analysis of GAN and GAN based inpainting methods.We firstly summarize and classify the recent representative inpainting algorithms. In addition,we discuss the characteristics and limitationsof these inpainting methods,and then organize and analyze the common evaluation indicators and public dataset. Finally,we describe thechallenges of image restoration and prospect the future development direction of image restoration technology.

相似文献/References:

[1]杨勇 汪继文.基于求解泊松方程和梯度的图像修复的研究[J].计算机技术与发展,2008,(02):98.
　YANG Yong,WANG Ji-wen.Research on Image Completion by Solving Poisson Equation and Gradient[J].,2008,(02):98.
[2]汪继文 林胜华 沈玉峰 邱剑锋.一种基于各向异性扩散的图像处理方法[J].计算机技术与发展,2008,(08):98.
　WANG Ji-wen,LIN Sheng-hua,SHEN Yu-feng,et al.An Approach for Image Restoration Based on Anisotropic Diffusion[J].,2008,(02):98.
[3]沈玉峰 汪继文 林胜华.一种大目标去除的图像修复方法[J].计算机技术与发展,2008,(11):159.
　SHEN Yu-feng,WANG Ji-wen,LIN Sheng-hua.A Method of Image Completion for Moving a Large Object[J].,2008,(02):159.
[4]林胜华 汪继文 沈玉峰 张义超.基于图像结构和纹理的快速修复方法[J].计算机技术与发展,2008,(12):88.
　LIN Sheng-hua,WANG Ji-wen,SHEN Yu-feng,et al.Fast Inpainting Approach Based on Image Structure and Texture[J].,2008,(02):88.
[5]胡超 吴骅 屈磊 韦穗.一种改进的快速步进图像修复算法[J].计算机技术与发展,2007,(05):206.
　HU Chao,WU Hua,QU Lei,et al.An Image Inpainting Algorithm Based on Improved Fast Marching Method[J].,2007,(02):206.
[6]汪潇 汪继文.基于水平线插值的图像修复算法[J].计算机技术与发展,2007,(09):61.
　WANG Xiao,WANG Ji-wen.Digital Image Impainting Using Level Line Interpolation[J].,2007,(02):61.
[7]段汉根 汪继文.基于邻域滤波的图像修复[J].计算机技术与发展,2007,(10):34.
　DUAN Han-gen,WANG Ji-wen.Image Inpainting via Neighborhood Filtering[J].,2007,(02):34.
[8]魏欣 蒋华伟.基于纹理和结构的图像修复算法研究[J].计算机技术与发展,2010,(09):90.
　WEI Xin,JIANG Hua-wei.Study on Inpainting Algorithm Based on Image Structure and Texture[J].,2010,(02):90.
[9]丁西明 段汉根 吴长勤.一种简单的块匹配图像修复算法[J].计算机技术与发展,2010,(11):80.
　DING Xi-ming,DUAN Han-gen,WU Chang-qin.Image Inpainting Based on Simple Block Matching[J].,2010,(02):80.
[10]张孟资,张立强.交替方向法在小波域图像修复中的应用[J].计算机技术与发展,2013,(10):235.
　ZHANG Meng-zi[],ZHANG Li-qiang[].Application of Alternating Direction Method in Wavelet Domain Image Inpainting[J].,2013,(02):235.

更新日期/Last Update: 2022-02-10