基于注意力机制的多尺度特征融合图像去雨方法
作者:
作者单位:

1.南京信息工程大学 人工智能学院未来技术学院;2.南京信息工程大学电子与信息工程学院;3.南京信息工程大学;4.杭州电子科技大学自动化学院

中图分类号:

TP391.41?????????????

基金项目:

国家自然科学基金项目(面上项目,重点项目,重大项目)


Multi-scale feature fusion image rain removal algorithm based on attention mechanism
Author:
Affiliation:

1.School of artificial intelligence (School of future technology),Nanjing University of Information Science and Technology;2.School of electronic and information engineering, Nanjing University of Information Science and Technology;3.School of electronic and information engineering, Nanjing University of Information Science and Technology,;4.School of Automation, Hangzhou University of Electronic Science and Technology

Fund Project:

The National Natural Science Foundation of China (General Program, Key Program, Major Research Plan)

  • 摘要
  • | |
  • 访问统计
  • |
  • 参考文献
  • | |
  • 引证文献
  • | |
  • 文章评论
    摘要:

    由于雨纹的分布和形状具有多样性,现有的去雨算法在去雨的同时也会产生图像背景模糊,泛化性能差等问题。针对这些问题,提出一种基于注意力机制的多尺度特征融合图像去雨方法。特征提取阶段由多个包含两个多尺度注意力残差块的残差组构成,多尺度注意力残差块利用多尺度特征提取模块提取及聚合不同尺度的特征信息,并通过坐标注意力进一步提高网络的特征提取能力。在组内进行局部特征融合,组间利用全局特征融合注意力模块更好的融合不同层次的特征,通过像素注意力使网络重点关注于雨纹区域。在仿真和真实雨像数据集上与其他现有的图像去雨算法相比,所提方法的定量指标有着明显提高,去雨后的图像视觉效果较好且具有良好的泛化性。

    Abstract:

    Due to the diversity of the distribution and shape of rain streaks, existing rain removal algorithms produce problems such as blurred image background and poor generalization performance while removing rain. A multi-scale feature fusion image rain removal method based on an attention mechanism is proposed to address these problems. The feature extraction stage consists of multiple residual groups containing two multi-scale attention residual blocks. The multi-scale attention residual blocks use the multi-scale feature extraction module to extract and aggregate feature information at different scales and further improve the feature extraction capability of the network through coordinate attention. Local feature fusion is performed within groups, and the global feature fusion attention module is used between groups to better fuse features at different levels and to focus the network on rain streaks regions through pixel attention. The quantitative metrics of the proposed method are significantly improved compared with other existing image rain removal algorithms on both simulated and real rain image datasets, and the visual effects of the rain removal images are better and have good generalization.

    参考文献
    [1] Kang L W,Lin C W,Fu Y H.Automatic single-image-based rain streaks removal via image decomposition [J]. IEEE transactions on image processing.2011,21(4);1742-55.
    [2] Li Y,Tan R T,Guo X J,et al.Rain streak removal using layer priors [C]//Proceedings of the IEEE Conference on Computer Vision and Pattern Recognition,Las Vegas,NV,USA:IEEE,2016:2736-2744.
    [3] Fu X Y,Huang J B,Ding X H,et al.Clearing the Skies: A Deep Network Architecture for Single-Image Rain Removal[J].IEEE Transactions on Image Processing,2017,26,(6):2944-2956.
    [4] Ren D W,Zuo W M,Hu Q H,et al.Progressive Image Deraining Networks: A Better and Simpler Baseline [C]. Proceedings of 2019 IEEE/CVF Conference on Computer Vision and Pattern Recognition,Long Beach,CA,USA:IEEE2019:3932-3941.
    [5] Fu X Y,Liang B R,Huang Y,et al.Lightweight Pyramid Networks for Image Deraining [J]. IEEE Transactions on Neural Networks and Learning Systems,2020,31,(6):1794-180.
    [6] Wang C,Xing X Y,Su Z X,et al.DCSFN:Deep cross-scale fusion network for single image rain removal [C]. Proceedings of the ACM International Conference on Multimedia,Seattle,USA:ACM,2020:1643-1651.
    [7] Yi Q S,Li J C,Dai Q,et al.Structure-Preserving Deraining with Residue Channel Prior Guidance[C]//2021 IEEE/CVF International Conference on Computer Vision (ICCV)(2021): 4218-4227.
    [8] Hou Q B,Zhou D Q,Feng J S.Coordinate Attention for Efficient Mobile Network Design[C]//Proceedings of 2021 IEEE/CVF Conference on Computer Vision and Pattern Recognition.2021:13713-13722.
    [9] He K M,Zhang X Y,Ren S Q,et al.Identity Mappings in Deep Residual Networks[C].European Conference on Computer Vision.2016:630-645.
    [10] 康健,管海燕,于永涛,景庄伟,刘超,高俊勇.基于RFA-LinkNet模型的高分遥感影像水体提取[J/OL].南京信息工程大学学报(自然科学版):1-14[2022-06-11].http://kns.cnki.net/kcms/detail/32.1801.N.20220120.1527.002.html
    [11] 胡序洋,高尚兵,汪长春,胡立伟,李少凡.LaneSegNet:一种高效的车道线检测方法[J/OL].南京信息工程大学学报(自然科学版):1-10[2022-06-11].http://kns.cnki.net/kcms/detail/32.1801.N.20211123.1544.002.html
    [12] 张凯,郭剑黎,胡军星,任俊霞,谭磊.基于注意力残差网络的WiFi设备的射频指纹识别[J/OL].南京信息工程大学学报(自然科学版):1-9[2022-06-11].http://kns.cnki.net/kcms/detail/32.1801.N.20210830.1652.002.html.
    [13] Xu Q,Wang Z L,Bai Y C,et al.FFA-Net:Feature fusion attention network for single image dehazing[C]//Proceedings of the AAAI Conference on Artificial Intelligence.2020:11908-11915.
    [14] 马婧婧,黄煜峰,陈翔.多尺度沙漏结构的单幅图像去雨算法研究[J].小型微型计算机系统,2021,42(03):561-565.
    [15] Yang W H,Tan R T,Feng J S,et al.Deep Joint Rain Detection and Removal from a Single Image [C]//2017 IEEE Conference on Computer Vision and Pattern Recognition,2017:1685-1694.
    [16] Cai L W,Li S W,Ren D W,et al.Dual recursive network for fast image deraining[C]//Proceedings of the IEEE International Conference on Image Processing.2019:2756-2760.
    [17] Yang Y Z,Lu H.Single image deraining via recurrent hierarchy enhancement network[C]//Proc of the 27th ACM International Conference on Multimedia.2019:1814-1822.
    [18] Shang W,Zhu P F,Ren D W,et al.Bilateral recurrent network for single image deraining[C]//International Conference on Acoustics,Speech and Signal Processing.2020:2503-250.
    相似文献
    引证文献
    引证文献 [0]
    网友评论
    网友评论
    分享到微博
    发 布
引用本文

刘忠洋,周杰,陆加新,缪则林,邵根富,江凯强,高伟.基于注意力机制的多尺度特征融合图像去雨方法[J].南京信息工程大学学报,,():

复制
分享
文章指标
  • 点击次数:865
  • 下载次数: 0
  • HTML阅读次数: 0
  • 引用次数: 0
历史
  • 收稿日期:2022-07-18
  • 最后修改日期:2022-10-09
  • 录用日期:2022-10-11

地址:江苏省南京市宁六路219号    邮编:210044

联系电话:025-58731025    E-mail:nxdxb@nuist.edu.cn

南京信息工程大学学报 ® 2025 版权所有  技术支持:北京勤云科技发展有限公司