摘要
图像超分辨率重建领域随着深度学习近年来的蓬勃发展得到更加广泛的应用,这种应用为图像超分辨率重建领域效果带来了显著的提升。文章的主要工作内容是对现有的一些基于深度学习的图像超分辨率重建方法进行分析和比较,对其中具有代表性的模型的相关算法进行了简要的介绍,并在当前公开的数据集上,对比分析了近年来较为主流和经典的图像超分辨率重建方法的实验结果。最后提出图像超分辨率重建技术面临的挑战,展望了该领域的未来发展方向。
With the rapid development of deep learning in recent years, the field of image super-resolution reconstruction has been more widely applied, which has brought significant improvement to the effect of image super-resolution reconstruction.This article’s main job content is to some existing image super-resolution reconstruction based on the deep learning method are analyzed and compared, some representative models of related algorithm was introduced in brief, and on the current public data sets, the comparison and analysis in recent years, more mainstream and classic image super-resolution reconstruction method of the experimental results. Finally, the challenges of image super-resolution reconstruction technology are presented, and the future development direction of this field is prospected.
作者
赵洋
Zhao Yang(College of Aeronautics and Astronautics,Heilongjiang Branch of National Computer Network Emergency Technical Processing and Coordination Center,Harbin 150000,China)
出处
《信息通信》
2020年第12期8-12,17,共6页
Information & Communications
关键词
深度学习
图像超分辨率重建
卷积神经网络
数据集
评价指标
deep learning
image super-resolution reconstruction
convolutional neural network
data set
evaluation index