Evaluation of Deep Convolutional Neural Network with Residual Learning for Remote Sensing Image Super Resolution

Rika Sustika

doi:10.18495/comengapp.v10i1.351

[1]R. Fernandez-beltran, P. Latorre-carmona, and F. Pla, “Single-Frame Super-Resolution in Remote Sensing : a Practical Overview,” International Journal of Remote Sensing, vol. 38, no. 1, pp. 314–354, 2017.
[2]C. Dong, C. C. Loy, K. He, and X. Tang, “Learning a Deep Convolutional Network for Image Super-Resolution,” in Computer Vision –European Conference on Computer Vision 2014, 2014, pp. 184–199.
[3]C. Dong, C. C. Loy, and K. He, “Image Super-Resolution Using Deep Convolutional Networks,” IEEE Transactions on Pattern Analysis and Machine Intelligence, vol. 38, no. 2, pp. 295–307, 2016.
[4]C. Dong, C. C. Loy, and X. Tang, “Accelerating the Super-Resolution Convolutional Neural Network,” in European Conference on Computer Vision (ECCV), 2016, pp. 391–407.
[5]J. Kim, J. K. Lee, and K. M. Lee, “Accurate Image Super-Resolution Using Very Deep Convolutional Networks,” in IEEE Conference on Computer Vision and Pattern Recognition (CVPR), 2016, pp. 1646–1654.

Abstract viewed - 1059 times
PDF downloaded - 886 times

Affiliations

Rika Sustika
Research Center for Informatics, LIPI

Content Top

Evaluation of Deep Convolutional Neural Network with Residual Learning for Remote Sensing Image Super Resolution

Rika Sustika

Vol 10 No 1 (2021)

DOI 10.18495/comengapp.v10i1.351

Submitted: Dec 8, 2020

Published: Feb 1, 2021

Abstract

Remote sensing images generally have low spatial resolution because of the limitations of sensing devices, bandwidth transmission, or storage capacity. An effective way to improve spatial resolution with low cost is by using algorithm based approach, known as super resolution (SR). In recent years, deep learning is super resolution technique that received special attention because it gave better performance than traditional method. In this research, we evaluated two simple deep learning architectures and explored parameters setting of deep convolutional neural network with residual learning, to achieve the trade-off between performance and speed or computational complexity, for implementation on remote sensing image super resolution. Results from the experiment show that deeper network with smaller number of filter gives faster model than shallow network with bigger number of filter, without sacrificing the performance.