Multispectral and Hyperspectral Image Fusion Based on Regularized Coupled Non-Negative Block-Term Tensor Decomposition

Abstract

The problem of multispectral and hyperspectral image fusion (MHF) is to reconstruct images by fusing the spatial information of multispectral images and the spectral information of hyperspectral images. Focusing on the problem that the hyperspectral canonical polyadic decomposition model and the Tucker model cannot introduce the physical interpretation of the latent factors into the framework, it is difficult to use the known properties and abundance of endmembers to generate high-quality fusion images. This paper proposes a new fusion algorithm. In this paper, a coupled non-negative block-term tensor model is used to estimate the ideal high spatial resolution hyperspectral images, its sparsity is characterized by adding 1-norm, and total variation (TV) is introduced to describe piecewise smoothness. Secondly, the different operators in two directions are defined and introduced to characterize their piecewise smoothness. Finally, the proximal alternating optimization (PAO) algorithm and the alternating multiplier method (ADMM) are used to iteratively solve the model. Experiments on two standard datasets and two local datasets show that the performance of this method is better than the state-of-the-art methods.

Qualitative and quantitative analysis of artificial neural network-based post-classification comparison to detect the earth surface variations using hyperspectral and multispectral datasets

“…The main reason behind such outcomes is that due to the low spatial resolution of the hyperspectral dataset, the mixed pixel interference causes the number of classes in the observed spectral response of the pixel. However, some of the attempts have also been made in the literature, such as the fusion of hyperspectral with high-resolution multispectral or panchromatic bands 71 , 72 . But this does not necessarily guarantee an enhancement in the spatial resolution and there are a lot of challenges are associate with the integration of the spatial and spectral information 73 , 74 .…”

“…However, some of the attempts have also been made in the literature, such as the fusion of hyperspectral with high-resolution multispectral or panchromatic bands. 71,72 But this does not necessarily guarantee an enhancement in the spatial resolution and there are a lot of challenges are associate with the integration of the spatial and spectral information. 73,74 Therefore, the current framework may be challenging for the detection of class categories with…”

Crossed Dual-Branch U-Net for Hyperspectral Image Super-Resolution

“…For example, Jin et al [38] presented a tensor network by fusing the high-order tensors that correspond to LRHS and HRMS images, designing a new regularization term named weightedgraph regularization. In response to the noise and non-smooth problems, Guo et al [39] inserted two different operators to design a tensor decomposition network. Based on tensor ring decomposition, He et al [40] designed a model that iteratively obtain corresponding core tensors from LRHS and HRMS images.…”

Nonnegative Block-Term Decomposition with the β-Divergence: Joint Data Fusion and Blind Spectral Unmixing

“…Therefore, in the absence of reference endmembers, R is estimated based on e.g., subspace identification [32], [33] Then, the L r are chosen to be large while satisfying the conditions of Theorem 2.2. Several works considered mixed-norm regularization to estimate the L r for various LL1-BTD, but these methods were not guaranteed to provide a unique solution when applied to the unmixing problem [13], [15]. Conversely, in [14] were given guarantees for unique recovery of the SRI and its mixing factors.…”

“…Previous works of the authors [14] used the block-term decomposition for joint fusion and unmixing in the presence of spectral variability. A recent work used the block-term decomposition for super-resolution only [15]. This decomposition was also successfully used to perform unmixing [16] on the SRI directly.…”