Generative Adversarial Network-Based Edge-Preserving Superresolution Reconstruction of Infrared Images

• Published in: International Journal of Digital Multimedia Broadcasting Vol. 2021; pp. 1 - 12
• Main Authors: Zhao, Yuqing, Fu, Guangyuan, Wang, Hongqiao, Zhang, Shaolei, Yue, Min
• Format: Journal Article
• Language: English
• Published: New York Hindawi 21.07.2021; Hindawi Limited; Wiley
• Subjects: Artificial neural networks; Generative adversarial networks; Image contrast; Image reconstruction; Infrared imagery; Infrared imaging systems; Magnetic resonance imaging; Mean square errors; Military electronics industry; Neural networks; Quality assessment; Visual effects; United States
• ISSN: 1687-7578; 1687-7586
• DOI: 10.1155/2021/5519508

The convolutional neural network has achieved good results in the superresolution reconstruction of single-frame images. However, due to the shortcomings of infrared images such as lack of details, poor contrast, and blurred edges, superresolution reconstruction of infrared images that preserves the edge structure and better visual quality is still challenging. Aiming at the problems of low resolution and unclear edges of infrared images, this work proposes a two-stage generative adversarial network model to reconstruct realistic superresolution images from four times downsampled infrared images. In the first stage of the generative adversarial network, it focuses on recovering the overall contour information of the image to obtain clear image edges; the second stage of the generative adversarial network focuses on recovering the detailed feature information of the image and has a stronger ability to express details. The infrared image superresolution reconstruction method proposed in this work has highly realistic visual effects and good objective quality evaluation results.