Image Compressed Sensing Using Convolutional Neural Network

被引:246
|
作者
Shi, Wuzhen [1 ,2 ]
Jiang, Feng [1 ,2 ]
Liu, Shaohui [1 ,2 ]
Zhao, Debin [1 ,2 ]
机构
[1] Harbin Inst Technol, Sch Comp Sci & Technol, Harbin 150001, Heilongjiang, Peoples R China
[2] Peng Cheng Lab, Shenzhen 518055, Guangdong, Peoples R China
基金
美国国家科学基金会;
关键词
Image reconstruction; Reconstruction algorithms; Matching pursuit algorithms; Deep learning; Computational complexity; Hardware; Compressed sensing; deep learning; convolutional neural network; sampling matrix; image reconstruction; SIGNAL RECOVERY; RECONSTRUCTION; BINARY;
D O I
10.1109/TIP.2019.2928136
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
In the study of compressed sensing (CS), the two main challenges are the design of sampling matrix and the development of reconstruction method. On the one hand, the usually used random sampling matrices (e.g., GRM) are signal independent, which ignore the characteristics of the signal. On the other hand, the state-of-the-art image CS methods (e.g., GSR and MH) achieve quite good performance, however with much higher computational complexity. To deal with the two challenges, we propose an image CS framework using convolutional neural network (dubbed CSNet) that includes a sampling network and a reconstruction network, which are optimized jointly. The sampling network adaptively learns the sampling matrix from the training images, which makes the CS measurements retain more image structural information for better reconstruction. Specifically, three types of sampling matrices are learned, i.e., floating-point matrix, & x007B;0, 1 & x007D;-binary matrix, and & x007B;-1, & x002B;1 & x007D;-bipolar matrix. The last two matrices are specially designed for easy storage and hardware implementation. The reconstruction network, which contains a linear initial reconstruction network and a non-linear deep reconstruction network, learns an end-to-end mapping between the CS measurements and the reconstructed images. Experimental results demonstrate that CSNet offers state-of-the-art reconstruction quality, while achieving fast running speed. In addition, CSNet with & x007B;0, 1 & x007D;-binary matrix, and & x007B;-1, & x002B;1 & x007D;-bipolar matrix gets comparable performance with the existing deep learning-based CS methods, outperforms the traditional CS methods. Experimental results further suggest that the learned sampling matrices can improve the traditional image CS reconstruction methods significantly.
引用
收藏
页码:375 / 388
页数:14
相关论文
共 50 条
  • [21] Image Denoising using Convolutional Neural Network
    Mehmood, Asif
    PATTERN RECOGNITION AND TRACKING XXXI, 2020, 11400
  • [22] Image enhancement using convolutional neural network
    Zhou, Abel
    Tan, Qi
    Davidson, Rob
    2020 INTERNATIONAL CONFERENCE ON IMAGE, VIDEO PROCESSING AND ARTIFICIAL INTELLIGENCE, 2020, 11584
  • [23] Remote Sensing Image Fusion With Deep Convolutional Neural Network
    Shao, Zhenfeng
    Cai, Jiajun
    IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING, 2018, 11 (05) : 1656 - 1669
  • [24] Improved convolutional neural network in remote sensing image classification
    Binghui Xu
    Neural Computing and Applications, 2021, 33 : 8169 - 8180
  • [25] Improved convolutional neural network in remote sensing image classification
    Xu, Binghui
    NEURAL COMPUTING & APPLICATIONS, 2021, 33 (14): : 8169 - 8180
  • [26] A NOVEL FRAMEWORK TO INTEGRATE CONVOLUTIONAL NEURAL NETWORK WITH COMPRESSED SENSING FOR CELL DETECTION
    Xue, Yao
    Ray, Nilanjan
    Hugh, Judith
    Bigras, Gilbert
    2017 24TH IEEE INTERNATIONAL CONFERENCE ON IMAGE PROCESSING (ICIP), 2017, : 2319 - 2323
  • [27] Remote Sensing Image Fusion Using Multi-Scale Convolutional Neural Network
    Wei Shi
    ChaoBen Du
    BingBing Gao
    JiNing Yan
    Journal of the Indian Society of Remote Sensing, 2021, 49 : 1677 - 1687
  • [28] Remote Sensing Image Retrieval Using Convolutional Neural Network Features and Weighted Distance
    Ye, Famao
    Xiao, Hui
    Zhao, Xuqing
    Dong, Meng
    Luo, Wei
    Min, Weidong
    IEEE GEOSCIENCE AND REMOTE SENSING LETTERS, 2018, 15 (10) : 1535 - 1539
  • [29] Remote Sensing Image Fusion Using Multi-Scale Convolutional Neural Network
    Shi, Wei
    Du, ChaoBen
    Gao, BingBing
    Yan, JiNing
    JOURNAL OF THE INDIAN SOCIETY OF REMOTE SENSING, 2021, 49 (07) : 1677 - 1687
  • [30] Compressive Sensing Image Reconstruction Using Super-Resolution Convolutional Neural Network
    Huang, Lilian
    Zhu, Zhonghang
    2018 2ND INTERNATIONAL CONFERENCE ON DIGITAL SIGNAL PROCESSING (ICDSP 2018), 2018, : 80 - 83