Joint-Sparse-Blocks and Low-Rank Representation for Hyperspectral Unmixing

Huang, Jie; Huang, Ting‐Zhu; Deng, Liang-Jian; Zhao, Xi-Le

doi:10.1109/tgrs.2018.2873326

Cited by 87 publications

(79 citation statements)

References 63 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Similarly, we multiply both sides by S on the equation (18) and substitute equation (19) into equation (18), the abundance matrix S can be updated as:…”

Section: B Optimizationmentioning

confidence: 99%

“…Some of these focus on the endmember extraction from statistical and geometrical aspects, such as Pixel Purity Index [13], N-FINDR [14], alternating projected subgradients [15], Vertex Component Analysis [16], independent component analysis [17], and minimum-volume-based unmixing algorithms [18], etc. Other methods address the problem of abundance estimation under the assumption that the endmembers are available [19]. With the almost universal success of deep learning, there are also examples of deep neural network based hyperspectral unmixing methods [20]- [22].…”

Section: Introductionmentioning

confidence: 99%

“…In addition, to avoid the influence of noise, many norm-based robust NMF methods have been proposed. The L 2,1 norm is commonly integrated into sparse NMF to achieve robustness for pixel noise and outlier rejection since it is rotationally invariant [19], [29], [30]. Additionally, the L 1,2 norm is also effective for solving band noise problems [31], [32].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

Subspace Structure Regularized Nonnegative Matrix Factorization for Hyperspectral Unmixing

Zhou

Zhang

Wang

et al. 2020

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Hyperspectral unmixing is a crucial task for hyperspectral images (HSI) processing, which estimates the proportions of constituent materials of a mixed pixel. Usually, the mixed pixels can be approximated using a linear mixing model. Since each material only occurs in a few pixels in real HSI, sparse nonnegative matrix factorization (NMF) and its extensions are widely used as solutions. Some recent works assume that materials are distributed in certain structures, which can be added as constraints to sparse NMF model. However, they only consider the spatial distribution within a local neighborhood and define the distribution structure manually, while ignoring the real distribution of materials that is diverse in different images. In this paper, we propose a new unmixing method that learns a subspace structure from the original image and incorporate it into the sparse NMF framework to promote unmixing performance. Based on the self-representation property of data points lying in the same subspace, the learned subspace structure can indicate the global similar graph of pixels that represents the real distribution of materials. Then the similar graph is used as a robust global spatial prior which is expected to be maintained in the decomposed abundance matrix. The experiments conducted on both simulated and real-world HSI datasets demonstrate the superior performance of our proposed method.

show abstract

“…Similarly, we multiply both sides by S on the equation (18) and substitute equation (19) into equation (18), the abundance matrix S can be updated as:…”

Section: B Optimizationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Subspace Structure Regularized Nonnegative Matrix Factorization for Hyperspectral Unmixing

Zhou

Zhang

Wang

et al. 2020

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

show abstract

“…However, they require appropriate computation equipment or parameters to maintain their performance, which is not always guaranteed and efficient. LMM assumption-based algorithms have more clear conceptual meaning to easily capture endmember extraction and abundance estimation, owing to multiple priors of data matrices such as sparse [7], [8], low-rank [9], and geometric [10] properties, which have attracted considerable attention [2], [3].…”

Section: Introductionmentioning

confidence: 99%

“…end for 15: end function Therefore, we propose the SENMAV algorithm, which identifies the endmembers by simultaneously considering its spectral features and spatial contextual. The endmembers in our SENMAV framework (10) obtain their spectral information under the data simplex via a maximum simplex volume framework (8) and (9), and the spatial information of the endmembers is acquired by means of the spatial energy prior (7). It is worth mentioning that both terms (i.e., spatial energy prior and maximum simplex volume) have different data scales and make different contributions to select endmembers.…”

mentioning

confidence: 99%

Spatial-Spectral Hyperspectral Endmember Extraction Using a Spatial Energy Prior Constrained Maximum Simplex Volume Approach

Shen

Bao

2020

IEEE J. Sel. Top. Appl. Earth Observations Remote Sensing

View full text Add to dashboard Cite

Endmember extraction algorithms (EEAs) are among the most commonly discussed types of hyperspectral image processing in the past three decades. This article proposes a spatial energy prior constrained maximum simplex volume (SENMAV) approach for spatial-spectral endmember extraction of hyperspectral images. SENMAV investigates the spatial information from the perspective of the spatial energy prior of a Markov random field (MRF), which is used as a regularization term of the traditional maximum volume simplex model to simultaneously constrain the selection of the endmembers in both the spatial and spectral viewpoints. This article sheds new light on spatial-spectral-based EEAs, as SEN-MAV well balances the tradeoff between endmember extraction accuracy and spatial attribute requirements of endmembers. Based on the spectral angle distance and root-mean-square error, experimental results on both synthetic and real hyperspectral datasets indicate that the proposed approach significantly improves the endmember extraction performance over current state-of-the-art spatial-spectral-based EEAs.

show abstract