English

Pan-sharpening via High-pass Modification Convolutional Neural Network

Computer Vision and Pattern Recognition 2021-05-26 v1 Machine Learning Image and Video Processing

Abstract

Most existing deep learning-based pan-sharpening methods have several widely recognized issues, such as spectral distortion and insufficient spatial texture enhancement, we propose a novel pan-sharpening convolutional neural network based on a high-pass modification block. Different from existing methods, the proposed block is designed to learn the high-pass information, leading to enhance spatial information in each band of the multi-spectral-resolution images. To facilitate the generation of visually appealing pan-sharpened images, we propose a perceptual loss function and further optimize the model based on high-level features in the near-infrared space. Experiments demonstrate the superior performance of the proposed method compared to the state-of-the-art pan-sharpening methods, both quantitatively and qualitatively. The proposed model is open-sourced at https://github.com/jiaming-wang/HMB.

Keywords

Cite

@article{arxiv.2105.11576,
  title  = {Pan-sharpening via High-pass Modification Convolutional Neural Network},
  author = {Jiaming Wang and Zhenfeng Shao and Xiao Huang and Tao Lu and Ruiqian Zhang and Jiayi Ma},
  journal= {arXiv preprint arXiv:2105.11576},
  year   = {2021}
}

Comments

5 pages, 5 figures, accepted by the 28th IEEE International Conference on Image Processing (ICIP 2021)

R2 v1 2026-06-24T02:25:33.353Z