Open AccessMulti-channel Feature Extraction and Super-Resolution Reconstruction of Remote Sensing Images
Author(s) -
Zheng Zhang,
Jiabin Zhang,
Changan Liu
Publication year - 2021
Publication title -
journal of physics. conference series
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.21
H-Index - 85
eISSN - 1742-6596
pISSN - 1742-6588
DOI - 10.1088/1742-6596/2006/1/012031
Subject(s) - computer science , artificial intelligence , computer vision , remote sensing , channel (broadcasting) , feature (linguistics) , feature extraction , image resolution , iterative reconstruction , pattern recognition (psychology) , geography , computer network , linguistics , philosophy
Super-resolution reconstruction is an imaging method to improve image resolution. It refers to reconstruct a clear high-resolution image from a low-resolution image. High-resolution remote sensing images can provide more detailed information and higher density, but in the field of remote sensing, because of the limitation of the hardware and vast distances, the remote sensing images are fuzzy sometimes. To facilitate subsequent tasks, this paper proposes a multi-channel feature extraction generative adversarial remote sensing image reconstruction method. According to the characteristics of remote sensing image, a generator is designed, which adds Laplace operator to enhance the edge information of the image, and uses multi-channel feature extraction, which not only enhances the ability of feature extraction but also reduces the number of parameters. In this paper, the super-resolution reconstruction task is carried out based on the 2X magnification factor, and the experimental results are evaluated on SET5/14 and NWPU-RESISC45 dataset. The experimental results show that the images generated by this method have a higher detailed texture and better super-resolution reconstruction effect of remote sensing images.