Academic Journal of Computing & Information Science, 2023, 6(13); doi: 10.25236/AJCIS.2023.061310.
Chang Che
School of Science and Technology for Future Life, Tokyo Denki University, Tokyo, Japan
In the digital age, it is a common problem that the image quality is damaged, especially the resolution is reduced. Traditional Super-Resolution (SR) methods, such as interpolation and reconstruction techniques, have achieved certain results, but they are still insufficient in processing complex and high-definition images. In recent years, Deep Learning (DL) technology has provided a new solution for SR. In this paper, the DL-based SR method is deeply studied, and related models and technologies are introduced in detail, including Convolutional Neural Network (CNN) and Generative Adversarial Network (GAN), and its principle, structure and performance are analyzed. At the same time, the SR technology based on DL is deeply studied, including key steps such as data set preprocessing, model training and optimization, and evaluation methods. Through comparative analysis, this paper shows the superiority and effect of DL method in SR. Through the above research contents and methods, it is expected to provide some theoretical support and technical guidance for the development of SR field and promote the development of this field to a higher level.
Deep Learning; Super-Resolution; Convolutional Neural Network; Generative Adversarial Network
Chang Che. Image Super-Resolution Processing Method Based on Deep Learning. Academic Journal of Computing & Information Science (2023), Vol. 6, Issue 13: 62-66. https://doi.org/10.25236/AJCIS.2023.061310.
[1] Nan Fangzhe, Qian Yurong, Xing Yanni, et al. Review of single-image super-resolution reconstruction based on deep learning [J]. Computer Application Research, 2020(2):6.
[2] Tan Chengbing, Yao Hongliang, Zhan Lin. Image super-resolution reconstruction based on deep learning and sparse coding [J]. Computer applications and software, 2022, 39(12):219-226.
[3] Zhang Han. Application of image super-resolution reconstruction based on deep learning [J]. Intelligent Computers and Applications, 2020, 10(5):4.
[4] Xing Suxiao, Chen Jinling, Li Xichao, et al. Overview of single-image super-resolution reconstruction based on deep learning [J]. Computer System Application, 2022(007):031.
[5] Jian Huang, Zhao Yuanyuan, Guo Ping, et al. Overview of super-resolution reconstruction methods of single image by deep learning [J]. Computer Engineering and Application, 2021, 57(18):11.
[6] Fan Mingming, Chi Yuan, Zhang Mingjin, Li Yunsong. Super-resolution reconstruction of chip images based on deep learning [J]. Pattern Recognition and Artificial Intelligence, 2019, 32(4): 353-360.
[7] Xia Hao, Lv Hongfeng, Luo Jun, et al. Research and application progress of image super-resolution deep learning [J]. Computer Engineering and Application, 2021, 57(24):51-60.
[8] Li Hong'an, Zheng Qiaoxue, Tao Ruolin, et al. Overview of image super-resolution research based on deep learning [J]. Acta Graphics, 2023, 44(1):15.
[9] Liu Yongxin, Duan Tiantian. Research on image super-resolution reconstruction technology based on deep learning [J]. Science and Technology and Innovation, 2018(23):4.
[10] Liu Cun, Li Yuanxiang, Zhou Yongjun, et al. Super-resolution reconstruction method of video image based on convolutional neural network [J]. Computer Application Research, 2019, 36(4):6.
[11] Li Jiaxing, Zhao Yongxian, Wang Jinghua. Overview of single image super-resolution reconstruction algorithm based on deep learning [J]. Acta automatica sinica, 2021, 47(10):23.