Super Resolution for Remote Sensing Images via Improved Residual Network

According to the processing characteristics of remote sensing image super-resolution, this paper studies a super resolution method based on improved residual network. First, we optimize the structure of the residual block to meet the needs of super-resolution tasks; then, we further deepen the netwo...

Full description

Saved in:
Bibliographic Details
Published in2020 5th International Conference on Mechanical, Control and Computer Engineering (ICMCCE) pp. 2295 - 2298
Main Authors Xie, Haiping, Jiang, Haiyang, Liu, Xiangyu, Li, Gaoyuan, Yang, Haitao
Format Conference Proceeding
LanguageEnglish
Published IEEE 01.12.2020
Subjects
component
Image reconstruction
Indexes
Remote sensing
residual networks
Sensors
super resolution
Superresolution
Task analysis
Visual effects
Online AccessGet full text

Cover

More Information
Summary:According to the processing characteristics of remote sensing image super-resolution, this paper studies a super resolution method based on improved residual network. First, we optimize the structure of the residual block to meet the needs of super-resolution tasks; then, we further deepen the network level, so that the network has a stronger learning ability. The reconstruction results on the remote sensing images dataset show that the improved residual network achieves better visual effect, and the objective evaluation index is significantly improved, which proves the effectiveness of the proposed method.
DOI:10.1109/ICMCCE51767.2020.00496