Hyperspectral image fusion based on sparse constraint NMF

Chen, Quan; Shi, Zhenwei; An, Zhenyu

doi:10.1016/j.ijleo.2013.07.061

Cited by 16 publications

(10 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…For example, with the good performance in blind source separation (BSS), NMF is frequently applied in HSI unmixing, where the endmembers are included in the basis matrix of NMF, while the corresponding abundances are included in coefficient matrix [18], [19]. NMF can also be employed in HSI fusion, e.g., sparse constraint NMF (SCNMF) was utilized for HSI fusion with panchromatic image in [20]. In details, unmixing is firstly performed, producing endmember-matrix and an abundance-matrix, then abundance-matrix is sharpened with the panchromatic image, and finally, the fused HSI is generated by solving the spectral constraint optimization problem.…”

Section: A Nmf and Gnmfmentioning

confidence: 99%

Semisupervised Dual-Dictionary Learning for Heterogeneous Transfer Learning on Cross-Scene Hyperspectral Images

Chen¹,

Ye²,

Ling³

et al. 2020

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

View full text Add to dashboard Cite

Lack of labeled training samples is a big challenge for hyperspectral image (HSI) classification. In recent years, cross-scene classification has become a new research topic. In cross-scene classification, two closely related HSI scenes are considered, one contains adequate labeled samples, namely source scene, while the other one contains only a few labeled samples, namely target scene. The goal of cross-scene classification is utilizing the labeled samples in source scene to benefit the classification in target scene. In most cases, different HSIs are imaged by different sensors, leading to different feature dimensions (numbers of bands) in different scenes. In this situation, heterogeneous transfer learning is demanded. In this work, we propose a heterogeneous transfer learning algorithm namely semisupervised dual-dictionary nonnegative matrix factorization (SS-DDNMF). SS-DDNMF consists of two contributions. 1) Dualdictionary nonnegative matrix factorization (DDNMF): DDNMF trains two dictionaries for source and target scenes, respectively, aiming at projecting the source and target features to a shared low-dimensional subspace, eliminating the difference between feature spaces. In DDNMF, within-scene and cross-scene graphs are built to maintain the similarities between pixels. 2) Semisupervised learning for target scene: as the limited number of labeled pixels in target scene will affect the graph building of DDNMF, semisupervised learning is adopted in target scene. In details, superpixel segmentation is adopted to generate pseudo labels for some unlabeled pixels, thus more "labeled" pixels can be considered for building better graphs. The effectiveness of SS-DDNMF is verified by experiments on cross-scene HSIs.

show abstract

Section: A Nmf and Gnmfmentioning

confidence: 99%

Semisupervised Dual-Dictionary Learning for Heterogeneous Transfer Learning on Cross-Scene Hyperspectral Images

Chen¹,

Ye²,

Ling³

et al. 2020

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

View full text Add to dashboard Cite

show abstract

“…Yokoya et al [27] defined a method based on coupled non-negative matrix factorization (CNMF), which is characterized as a dual-loop structure that comprises updating and unmixing operation to generate cost functions with respect to spectral and spatial degradation convergence. Nevertheless, the practical application of CNMF should be proven because of the difficulty of involving and implementing numerous influencing factors [28].…”

Section: State Of the Artmentioning

confidence: 99%

Image Fusion in Remote Sensing Based on Spectral Unmixing and Improved Non - negative Matrix Factorization Algorithm

He¹,

Wang²,

Lai³

et al. 2018

JESTR

View full text Add to dashboard Cite

The spectral resolution of images improves with the continuous development of imaging technology in remote sensing. Traditional image fusion technologies result in problems in hyperspectral image fusion, such as spectral distortion and unclear physical meaning. A fast model based on improved non-negative matrix factorization (INMF) for hyperspectral and panchromatic image fusion was proposed to optimize the results of image fusion. First, the hyperspectral image was decomposed into endmember and abundance matrices using the INMF algorithm. The panchromatic image was then used to sharpen the abundance matrix. The spectral and sparse constraints were imposed on the objective function of the model. Finally, the resultant fused image was obtained by reconstructing the resolved endmember and abundance matrices. Results show that the proposed method is superior to other methods in terms of subjective assessment and objective analysis. As far as the indices of information entropy, correlation coefficient, Q-average, and spectral divergence are concerned, the proposed method surpasses those of second-best methods (Ehler, classical NMF, Ehler and Ehler) by 2.06%, 0.36%, 0.91%, and 56.31%, respectively, in the synthetic data experiment, and exceeds those of the second-best methods (Ehler, high-pass filtering, high-pass filtering and Ehler) by 0.13%, 10.05%, 3.89%, and 7.26%, respectively, in the real data experiment. Moreover, runtime is proportional to data size, and the proposed method takes the least time when image size is between 1 MB and 480 MB. This study provides theoretical reference for fast hyperspectral and panchromatic image fusion.

show abstract

“…Considering the low spatial resolution of hyperspectral image, obviously spectrum mixture existed, and a representative dictionary was hard to decompose directly. Incorporating sparse prior to the NMF model, Chen [36] proposed a sparse constraint nonnegative matrix factorization (SCNMF) and used the PAN image to sharpen the abundance matrix. By adding the sparse constraint in the fusion procedure, spectral information was better preserved along with the increase of computation complexity.…”

Section: Introductionmentioning

confidence: 99%

Hyperspectral Image Resolution Enhancement Approach Based on Local Adaptive Sparse Unmixing and Subpixel Calibration

Teng

Zhang

et al. 2018

Remote Sensing

View full text Add to dashboard Cite

Unmixing based fusion aims at generating a high spectral-spatial resolution image (HSS) with the same surface features of the high spatial resolution multispectral image (MS) and low spatial resolution hyperspectral image (HS). In this paper, a new fusion method is proposed to improve the fusion performance by taking further advantage of the distribution characteristics of ground objects. First, we put forward a local adaptive sparse unmixing based fusion (LASUF) algorithm, in which the sparsity of the abundance matrices is appended as the constraint to the optimization fusion, considering the limited categories of ground objects in a specific range and the local correlation of their distribution. Then, to correct the possible original subpixel misregistrations or those introduced by the fusion procedures, a subpixel calibration method based on optimal matching adaptive morphology filtering (OM-AMF) is designed. Experiments on various datasets captured by different sensors demonstrate that the proposed fusion algorithm surpasses other typical fusion techniques in both spatial and spectral domains. The proposed method effectively preserves the spectral composition features of the isolated ground objects within a small area. In addition, the OM-AMF postprocessing is able to spatially correct the fusion results at a subpixel level and preserve the spectral features simultaneously.

show abstract

Hyperspectral image fusion based on sparse constraint NMF

Cited by 16 publications

References 16 publications

Semisupervised Dual-Dictionary Learning for Heterogeneous Transfer Learning on Cross-Scene Hyperspectral Images

Semisupervised Dual-Dictionary Learning for Heterogeneous Transfer Learning on Cross-Scene Hyperspectral Images

Image Fusion in Remote Sensing Based on Spectral Unmixing and Improved Non - negative Matrix Factorization Algorithm

Hyperspectral Image Resolution Enhancement Approach Based on Local Adaptive Sparse Unmixing and Subpixel Calibration

Contact Info

Product

Resources

About