• 中国计算机学会会刊
  • 中国科技核心期刊
  • 中文核心期刊

计算机工程与科学 ›› 2021, Vol. 43 ›› Issue (01): 112-124.

• 图形与图像 • 上一篇    下一篇

基于深度学习的单幅图像超分辨率重建综述

李彬1,喻夏琼2,王平1,傅瑞罡1,张虹3   

  1. (1.国防科技大学电子科学学院,湖南 长沙 410073;2.32021部队,北京 100094;

    3.中国船舶科学研究中心,江苏 无锡 214000)

  • 收稿日期:2019-12-17 修回日期:2020-03-28 接受日期:2021-01-25 出版日期:2021-01-25 发布日期:2021-01-22

A survey of single image super-resolution reconstruction based on deep learning

LI Bin1,YU Xia-qiong2,WANG Ping1,FU Rui-gang1,ZHANG Hong3   

  1. (1.College of Electronic Science and Technology,National University of Defense Technology,Changsha 410073;

    2.32021 Troops of the PLA,Beijing 100094;3.China Ship Science Research Center,Wuxi 214000,China)

  • Received:2019-12-17 Revised:2020-03-28 Accepted:2021-01-25 Online:2021-01-25 Published:2021-01-22

摘要: 单幅图像超分辨率SISR重建指从单幅低分辨率图像恢复出高分辨率图像。深度学习方法越来越多地用于图像超分辨重建领域,由于深度网络模型可以自主学习低分辨率图像到高分辨率图像之间的映射关系,与传统方法相比在该领域展现出了更好的重建效果,因而
基于深度学习的方法已经成为目前图像超分辨率重建领域的主流方向。围绕现有的超分辨深度网络模型在重建方式、结构组成和损失函数方面展开的探索进行了综合论述,通过比较不同模型之间存在的异同点,分析了不同的模型构建方法存在的优缺点及适应的应用场景,同时比较不同网络模型在主流测试数据集上的重建效果,并对该领域的未来研究方向进行了展望。


关键词: 深度学习, 超分辨率重建, 神经网络, 信息融合

Abstract: Single image super-resolution (SISR) refers to the recovery of a high-resolution image from a single low-resolution image. With deep learning used in the field of image super-resolution, deep networks can independently learn the mapping relationship between low-resolution and high-resolution training images, showing better reconstruction performance than the traditional methods. Therefore, deep learning has become dominant in super-resolution. This paper focuses on the exploration of the existing deep network model of super-resolution in terms of reconstruction mode, network structure, and loss function. By comparing the similarities and differences between different models, the advan- tages and disadvantages of different model building methods and the applicable application scenarios are analyzed. Meanwhile, the reconstruction results of different network models on the benchmark test datasets are compared and the potential directions are concluded.



Key words: deep learning, super-resolution reconstruction, neural network, information fusion