Journal Article

UPGAN: An Unsupervised Generative Adversarial Network Based on U-Shaped Structure for Pansharpening

Feng, Yuting, Chu, Xing, Wang, Qianqian, Miao, Shengfa, Zheng, Huangqimei, Jin, Xin, Jiang, Qian

2024

Overview

Pansharpening is the fusion of panchromatic images and multispectral images to obtain images with high spatial resolution and high spectral resolution, which have a wide range of applications. At present, methods based on deep learning can fit the nonlinear features of images and achieve excellent image quality; however, the images generated with supervised learning approaches lack real-world applicability. Therefore, in this study, we propose an unsupervised pansharpening method based on a generative adversarial network. Considering the fine tubular structures in remote sensing images, a dense connection attention module is designed based on dynamic snake convolution to recover the details of spatial information. In the stage of image fusion, the fusion of features in groups is applied through the cross-scale attention fusion module. Moreover, skip layers are implemented at different scales to integrate significant information, thus improving the objective index values and visual appearance. The loss function contains four constraints, allowing the model to be effectively trained without reference images. The experimental results demonstrate that the proposed method outperforms other widely accepted state-of-the-art methods on the QuickBird and WorldView2 data sets.

• Publisher: MDPI AG
• Subject: Computational linguistics / Computer vision / Decomposition / Deep learning / Generative adversarial networks / image fusion / Image quality / Information processing / Language processing / Methods / Modules / Multisensor fusion / Natural language interfaces / Neural networks / Optimization / pansharpening / Remote sensing / Satellites / Semantics / Sensors / Snakes / Spatial data / Spatial discrimination learning / Spatial resolution / spectral analysis / Spectral resolution / Supervised learning / Visual perception / Wavelet transforms