Residual Forward-Subtracted U-Shaped Network for Dynamic and Static Image Restoration

被引:2
作者
Jung, Ho Min [1 ]
Kim, Byeong Hak [2 ]
Kim, Min Young [1 ,3 ]
机构
[1] Kyungpook Natl Univ, Sch Elect Engn, Daegu 41566, South Korea
[2] Hanwha Syst Co, Dept Optron, Gumi 39376, South Korea
[3] Kyungpook Natl Univ, Res Ctr Neurosurg Robot Syst, Daegu 41566, South Korea
关键词
Image restoration; Task analysis; Machine learning algorithms; AWGN; Software algorithms; Noise reduction; Heuristic algorithms; Restoration; multi-type noises; image denoising; image enhancement; convolutional neural network; residual learning;
D O I
10.1109/ACCESS.2020.3011580
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Advanced image sensors with high resolution are now being developed for specially purposed electro-optical systems, with research focused on robust image quality performance in terms of super resolution and noise removal under various environmental conditions. Recently, machine-learning and deep-learning methods have been studied as the best practical techniques for restoration to improve the deteriorated image quality of sensors. However, these methods show limitations and side effects of image degradation such as image non-uniformity. In this paper, we analyze and randomly generate additive white Gaussian noise, non-uniform line noise, and dark saturation as representative image degradations. We then propose an advanced U-net model based on global and local residual learning in order to restore complexly deteriorated images. The proposed method shows unparalleled performance compared to alternative models and previous studies. In particular, various complex noise components are minimized and improved with equal quality so that variation between sequential images is minimized. These findings leverage mutual corroboration of quantitative and qualitative evaluation metrics. In the future, the proposed model is expected to contribute to a wide range of field applications such as defense, surveillance, and video media for image quality enhancement technologies.
引用
收藏
页码:145401 / 145412
页数:12
相关论文
共 41 条
  • [1] NTIRE 2017 Challenge on Single Image Super-Resolution: Dataset and Study
    Agustsson, Eirikur
    Timofte, Radu
    [J]. 2017 IEEE CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION WORKSHOPS (CVPRW), 2017, : 1122 - 1131
  • [2] [Anonymous], 2017, SCI REP UK, DOI DOI 10.1038/S41598-017-13595-7
  • [3] [Anonymous], 2020, IEEE ACCESS, DOI DOI 10.1109/ACCESS.2020.2966794
  • [4] [Anonymous], 2017, ISPRS J PHOTOGRAMM, DOI DOI 10.1016/J.ISPRSJPRS.2017.08.004
  • [5] [Anonymous], 2015, LECT NOTES COMPUT SC, DOI DOI 10.1007/978-3-319-24574-4_28
  • [6] Anwar S., 2019, ARXIV190407523
  • [7] The Perception-Distortion Tradeoff
    Blau, Yochai
    Michaeli, Tomer
    [J]. 2018 IEEE/CVF CONFERENCE ON COMPUTER VISION AND PATTERN RECOGNITION (CVPR), 2018, : 6228 - 6237
  • [8] Chang Y., 2016, IEEE T GEOSCI REMOTE, V54, P3235
  • [9] Infrared Aerothermal Nonuniform Correction via Deep Multiscale Residual Network
    Chang, Yi
    Yan, Luxin
    Liu, Li
    Fang, Houzhang
    Zhong, Sheng
    [J]. IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2019, 16 (07) : 1120 - 1124
  • [10] Transformed Low-rank Model for Line Pattern Noise Removal
    Chang, Yi
    Yan, Luxin
    Zhong, Sheng
    [J]. 2017 IEEE INTERNATIONAL CONFERENCE ON COMPUTER VISION (ICCV), 2017, : 1735 - 1743