Computer Engineering and Applications ›› 2019, Vol. 55 ›› Issue (6): 173-177.DOI: 10.3778/j.issn.1002-8331.1712-0048

Previous Articles     Next Articles

Motion Deblurring Based on Generative Adversarial Networks

SANG Liang1,2,3,4, GAO Shuang4, YIN Zengshan1,2,4   

  1. 1.Shanghai Institute of Microsystem and Information Technology, Chinese Academy of Sciences, Shanghai 200050, China
    2.University of Chinese Academy of Sciences, Beijing 100049, China
    3.School of Information Science and Technology, Shanghai Tech University, Shanghai 201210, China
    4.Shanghai Engineering Center for Microsatellites, Shanghai 201203, China
  • Online:2019-03-15 Published:2019-03-14


桑  亮1,2,3,4,高  爽4,尹增山1,2,4   

  1. 1.中国科学院 上海微系统与信息技术研究所,上海 200050
    2.中国科学院大学,北京 100049
    3.上海科技大学 信息科学与技术学院,上海 201210
    4.上海微小卫星工程中心,上海 201203

Abstract: Image motion blur is a very challenging problem caused by camera shaking or object movements. In order to tackle this problem, the paper proposes a deep convolutional neural network based on generative adversarial networks method. The proposed method can restore a clear image in an end-to-end way without estimating blur kernel. By introducing adversarial loss based on generative adversarial networks and modifying the residual network structure, the proposed method can restore image details effectively. Then this paper trains this deep convolutional neural network model on public datasets. Finally, it is proved that the proposed method achieves good results according to the test on blurry image benchmark datasets.

Key words: motion blur, image restoration, generative adversarial networks, deep learning

摘要: 针对相机成像时相机抖动、物体运动等导致图像产生运动模糊这一十分具有挑战性的问题,提出基于生成对抗网络的深度卷积神经网络来复原模糊图像的解决方案。该方案省略了模糊核估计的过程,采用端对端的方式直接获取复原图像;通过引入生成对抗网络思想的对抗损失和对残差网络进行改进,有效地复原了图像的细节信息。最后通过训练此深度卷积神经网络模型并在相关模糊复原基准数据集上测试,证明了该方案取得了较好的结果。

关键词: 运动模糊, 图像复原, 生成对抗网络, 深度学习