.png){kind=icon}
返回
A subspace-based total variation regularization for sparse hyperspectral image unmixing
.png){kind=icon}
PRE
AI
DOI：10.1080/01431161.2025.2522945.png)
.png){kind=icon}
原文求助.png){kind=icon}
分享.png){kind=icon}
收藏
摘要
En 中文
The objective of hyperspectral image unmixing is to determine the fractional abundances of constituent materials within mixed pixels. With large spectral libraries available, sparse spectral unmixing assumes that each mixed pixel consists of only a few endmembers, which leads to excessively large abundance dimensions obtained by linear model decomposition. This increased dimensionality complicates the extraction of meaningful information in sparse unmixing problems. To solve it, we propose a subspace-based total variation (TV) regularization for sparse hyperspectral image unmixing. Particularly, we first project the abundance matrix into a lower dimensional subspace by using a semi-orthogonal matrix and then exploit the spatial information in the projected abundance maps by adapting the TV regularization. This proposed approach not only eliminates unnecessary computations on redundant and nonactive abundance maps, but also enhances the extraction of spatial information from the projected abundance maps in the gradient domain. We propose an iterative algorithm based on the alternating direction multiplier method, calledabundance projected subspace low-rank unmixing with TV(APSLRU-TV). Simulation and real-world experimental results demonstrate that the proposed algorithm achieves high estimation accuracy and effectively captures spatial information.
Keyword:
Hyperspectral unmixing
projected matrix
total variation regularization
abundance estimation
期刊
IF:
2.6
论文数:
1.2W
被引数:
2.7W
机构
引用论文
.png){kind=icon}
Fully constrained least squares linear spectral mixture analysis method for material quantification in hyperspectral imagery高光谱图像物质量化的全约束最小二乘线性光谱混合分析方法
PREAI
分享收藏
Spatial Group Sparsity Regularized Nonnegative Matrix Factorization for Hyperspectral Unmixing用于高光谱分解的空间群稀疏正则化非负矩阵分解
PREAI
分享收藏Simultaneously Sparse and Low-Rank Abundance Matrix Estimation for Hyperspectral Image Unmixing高光谱图像分解的同时稀疏和低秩丰度矩阵估计
OAAI
OAAI分享收藏
查看更多内容