TY - GEN
T1 - Single Image Super-Resolution with Application to Remote-Sensing Image
AU - Deeba, Farah
AU - Dharejo, Fayaz Ali
AU - Zhou, Yuanchun
AU - Ghaffar, Abdul
AU - Memon, Mujahid Hussain
AU - Kun, She
N1 - Funding Information:
This work is supported by National Natural Science Foundation of China (NSFC) under the grant number 61836013 and National Key Research Development and Plan of China under the grant 2016YFB501901
Publisher Copyright:
© 2020 IEEE.
PY - 2020/10/6
Y1 - 2020/10/6
N2 - To improve the resolution of satellite images, many researchers are committed to machine learning and neural network-based SR methods. SR has multiple residual network frameworks in deep learning that have improved performance and can extend thousands of layers in the system. However, each layer improves accuracy by doubling the number of layers, although training thousands of layers are too expensive, the process is slow, and there are functional recovery issues. To address these issues, we propose a super-resolution wide remote sensing residual network (WRSR), in which we increase the width and reduce the depth of the residual network, due to decreasing the depth of the network our model reduced memory costs. To enhance the resolution of the single image we showed that our method improves training loss performance by performing the weight normalization instead of augmentation technology. The results of the experiment show that the method performs well in terms of quantitative indicators (PSNR) and (SSIM).
AB - To improve the resolution of satellite images, many researchers are committed to machine learning and neural network-based SR methods. SR has multiple residual network frameworks in deep learning that have improved performance and can extend thousands of layers in the system. However, each layer improves accuracy by doubling the number of layers, although training thousands of layers are too expensive, the process is slow, and there are functional recovery issues. To address these issues, we propose a super-resolution wide remote sensing residual network (WRSR), in which we increase the width and reduce the depth of the residual network, due to decreasing the depth of the network our model reduced memory costs. To enhance the resolution of the single image we showed that our method improves training loss performance by performing the weight normalization instead of augmentation technology. The results of the experiment show that the method performs well in terms of quantitative indicators (PSNR) and (SSIM).
KW - Low-resolution LR
KW - remote-sensing images
KW - super-resolution (SR)
KW - wide residual block
UR - http://www.scopus.com/inward/record.url?scp=85104408111&partnerID=8YFLogxK
U2 - 10.1109/GCWOT49901.2020.9391625
DO - 10.1109/GCWOT49901.2020.9391625
M3 - Conference contribution
AN - SCOPUS:85104408111
T3 - 2020 Global Conference on Wireless and Optical Technologies, GCWOT 2020
BT - 2020 Global Conference on Wireless and Optical Technologies, GCWOT 2020
PB - Institute of Electrical and Electronics Engineers Inc.
T2 - 2020 Global Conference on Wireless and Optical Technologies, GCWOT 2020
Y2 - 6 October 2020 through 8 October 2020
ER -