A New Sparse Subspace Clustering Algorithm for Hyperspectral Remote Sensing Imagery

Zhai, Han; Zhang, Hongyan; Zhang, Liangpei; Li, Pingxiang; Plaza, Antonio

doi:10.1109/lgrs.2016.2625200

Cited by 95 publications

(73 citation statements)

References 16 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In order to evaluate the effectiveness of the proposed KSSC-SMP algorithm, the following stateof-the-art clustering approaches were implemented for comparative purposes: FCM-S1 [8], clustering by fast search and find of density peaks (CFSFDP) [35], SSC [13], S 4 C [15], and L2-SSC [16]. The number of clusters was determined by carefully observing the original image, and the parameters of each clustering method were carefully optimized through the well-known grid search strategy.…”

Section: Resultsmentioning

confidence: 99%

“…The geometric resolution of this image is 1.3 m, with nine classes contained in the scene. Considering the computational complexity, a typical sub-set at a size of 200 × 100 was utilized as the test data [15][16][17], with eight main land-cover Remote Sens. 2017, 9, 335 9 of 15 classes: Metal sheet, Asphalt, Meadows, Trees, Bare soil, Bitumen, Bricks, and Shadows.…”

Section: Experimental Results Obtained With the University Of Pavia Imentioning

confidence: 99%

“…We then construct the similarity graph using the sparse coefficient matrix [15][16][17]. Meanwhile, the symmetric form is adopted to enhance the connectivity of the graph:…”

Section: Sparse Subspace Clustering (Ssc)mentioning

confidence: 99%

“…As a perfect combination of these two theories, the sparse subspace clustering (SSC) algorithm [13,14] has achieved great success in the computer vision field. By treating pixels from the same land-cover class that have different spectra as belonging to the same subspace, the SSC algorithm has also shown great potential in HSI clustering [15][16][17], which is mainly due to its capacity to address the spectral variability problem. However, SSC interprets the HSIs only with the spectral information, and it ignores the rich spatial information contained in the images.…”

Section: Introductionmentioning

confidence: 99%

“…In a recent study [15], the spectral-spatial sparse subspace clustering (S 4 C) algorithm was proposed to improve clustering performance by the use of a spectral weighting strategy to relieve the high correlation problem between hyperspectral pixels and a local averaging constraint to incorporate the spatial information contained in the eight-connected neighborhood. As an enhanced approach incorporating spatial information, the 2 -norm regularized sparse subspace clustering (L2-SSC) algorithm [16] was developed to exploit the inherent spectral-spatial properties of HSIs in a more elaborate way, and can achieve a good effect.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Kernel Sparse Subspace Clustering with a Spatial Max Pooling Operation for Hyperspectral Remote Sensing Data Interpretation

Zhai

Zhang

et al. 2017

Remote Sensing

Self Cite

View full text Add to dashboard Cite

Hyperspectral image (HSI) clustering is generally a challenging task because of the complex spectral-spatial structure. Based on the assumption that all the pixels are sampled from the union of subspaces, recent works have introduced a robust technique-the sparse subspace clustering (SSC) algorithm and its enhanced versions (SSC models incorporating spatial information)-to cluster HSIs, achieving excellent performances. However, these methods are all based on the linear representation model, which conflicts with the well-known nonlinear structure of HSIs and limits their performance to a large degree. In this paper, to overcome these obstacles, we present a new kernel sparse subspace clustering algorithm with a spatial max pooling operation (KSSC-SMP) for hyperspectral remote sensing data interpretation. The proposed approach maps the feature points into a much higher dimensional kernel space to extend the linear sparse subspace clustering model to nonlinear manifolds, which can better fit the complex nonlinear structure of HSIs. With the help of the kernel sparse representation, a more accurate representation coefficient matrix can be obtained. A spatial max pooling operation is utilized for the representation coefficients to generate more discriminant features by integrating the spatial-contextual information, which is essential for the accurate modeling of HSIs. This paper is an extension of our previous conference paper, and a number of enhancements are put forward. The proposed algorithm was evaluated on two well-known hyperspectral data sets-the Salinas image and the University of Pavia image-and the experimental results clearly indicate that the newly developed KSSC-SMP algorithm can obtain very competitive clustering results for HSIs, outperforming the current state-of-the-art clustering methods.

show abstract

Section: Resultsmentioning

confidence: 99%

Section: Experimental Results Obtained With the University Of Pavia Imentioning

confidence: 99%

“…We then construct the similarity graph using the sparse coefficient matrix [15][16][17]. Meanwhile, the symmetric form is adopted to enhance the connectivity of the graph:…”

Section: Sparse Subspace Clustering (Ssc)mentioning

confidence: 99%