Domain Transform Model Driven by Deep Learning for Anti-Noise Hyperspectral and Multispectral Image Fusion

While fusion of hyperspectral images with low spatial resolution and multispectral images with high spatial resolution has achieved significant success, high-quality fusion between noisy images has always been challenging. In this paper, We propose a domain transform model driven by deep learning for anti-noise hyperspectral and multi-spectral image fusion (DTAFN). This marks the first time that wavelet decomposition theory is combined with deep learning for noise reduction in hyperspectral and multispectral image fusion. DTAFN initially decomposes hyperspectral and multispectral images into frequency components and constructs a novel Feature Interaction Fusion Module. This module, while using multispectral images to guide the removal of noise from hyperspectral images, also achieves the fusion of spatial and spectral information. Furthermore, it maps the fused features to a lower-dimensional subspace to enhance computational efficiency. Additionally, we introduce a spatial-spectral self-attention mechanism to optimize the reconstructed frequency components using the subspace features. In the end, the wavelet inverse transform is used to reconstruct the clean fused image. It’s worth noting that the extraction of the subspace is considered a process of non-linear low-rank component extraction, which, to a certain extent, suppresses noise signals.. Numerous experiments of mixed noise image fusion are carried out, and the experimental results show that DTAFN can obtain high-quality fusion results, is robust, and superior to state-of-the-art methods.