Sparse unmixing of hyperspectral data with bandwise model

Li, Chang; Liu, Yu; Cheng, Juan; Song, Rencheng; Ma, Jiayi; Sui, Chenhong; Chen, Xun

doi:10.1016/j.ins.2019.10.036

Cited by 18 publications

(8 citation statements)

References 45 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Only in few cases, the HSIs are degraded by other types of noise such as impulse noise or deadlines, which can be classified as sparse noise [23], [27]. Therefore, the SU bandwise model [23], [26] is adopted in this paper, concerning about mixed noise cases, which can be expressed as…”

Section: B Bandwise Modelmentioning

confidence: 99%

“…The sparse noise term S has potential row sparsity that only a few bands of HSIs have impulse noise or deadlines [25]. Therefore, we impose l 0 norm constraints on the S [26], and further obtain the following objective function…”

Section: L 20 Norm Constraint Based On Sparse Noise Term Smentioning

confidence: 99%

“…In this section, we verify the unmixing performance of the proposed BMSPI on simulated and real data sets, and compare it with five advanced SU algorithms, such as CLSUnSAL [17], SUnSPI [18], CSUnL0 [31], JSpBLRU [14] and SUBM [26]. Since the number of regularization parameters set for all algorithms is no more than two, the optimal regularization parameters of all algorithms are chosen from the sequence {10 −4 , 10 −3 , 10 −2 , 10 −1 , 10 0 , 10 1 } in the follow-up experiments, according to the suggestions in their literature.…”

Section: A Comparison Algorithms and Evaluation Criteriamentioning

confidence: 99%

“…In order to compare the effects of the unmixing and noise reduction of various algorithms quantitatively, we use the rootmean-square error (RMSE) [26] and the mean peak signal-tonoise ratio (MPSNR) [23] to evaluate the unmixing performance of all algorithms.…”

Section: A Comparison Algorithms and Evaluation Criteriamentioning

confidence: 99%

“…Therefore, bandwise model of HSIs' noise was carried out, combined with low-rank matrix decomposition, to improve the suppression ability among different types of noise. Li et al [26] assumed that the Gaussian noise of each band was independently and uniformly distributed. By considering different types of noise under the maximum a posteriori probability, effectively solves the problem of mixed noise in SU.…”

Section: Introductionmentioning

confidence: 99%

See 4 more Smart Citations

Bandwise Model Based on Spectral Prior Information for Sparse Unmixing

Kang

Jiang

et al. 2021

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

View full text Add to dashboard Cite

The purpose of sparse unmixing (SU) is to find the optimal spectral subset from the spectral library and uses this subset to model each pixel in the hyperspectral data. The existing SU methods concern Gaussian noise a lot and focus less on the varied intensity of Gaussian noise in different bands and other types of noise, e.g., impulse noise and deadlines. Besides, the high coherence of the spectral library limits the performance of SU. Given the above problems, this paper proposes a new method, called Bandwise Model based on Spectral Prior Information (BMSPI). This proposed BMSPI models the Gaussian noise across different spectral bands and the other types of mixed noise under the maximum a posteriori probability framework, and decreases the effect of high coherence in the spectral library with the spectral prior information. The Alternating Direction Method of Multipliers (ADMM) is adopted to solve the BMSPI. The results of the simulated and real data experiments show that the bandwise model can suppress noises of different types effectively, and the spectral prior information is conducive to guide SU. The advantage of BMSPI is that the mentioned information is used completely. Thus, the accuracy of abundance estimation is improved.

show abstract