Endmember Learning with K-Means through SCD Model in Hyperspectral Scene Reconstructions

Abstract

This paper proposes a simple yet effective method for improving the efficiency of sparse coding dictionary learning (DL) with an implication of enhancing the ultimate usefulness of compressive sensing (CS) technology for practical applications, such as in hyperspectral imaging (HSI) scene reconstruction. CS is the technique which allows sparse signals to be decomposed into a sparse representation “a” of a dictionary D u . The goodness of the learnt dictionary has direct impacts on the quality of the end results, e.g., in the HSI scene reconstructions. This paper proposes the construction of a concise and comprehensive dictionary by using the cluster centres of the input dataset, and then a greedy approach is adopted to learn all elements within this dictionary. The proposed method consists of an unsupervised clustering algorithm (K-Means), and it is then coupled with an advanced sparse coding dictionary (SCD) method such as the basis pursuit algorithm (orthogonal matching pursuit, OMP) for the dictionary learning. The effectiveness of the proposed K-Means Sparse Coding Dictionary (KMSCD) is illustrated through the reconstructions of several publicly available HSI scenes. The results have shown that the proposed KMSCD achieves ~40% greater accuracy, 5 times faster convergence and is twice as robust as that of the classic Spare Coding Dictionary (C-SCD) method that adopts random sampling of data for the dictionary learning. Over the five data sets that have been employed in this study, it is seen that the proposed KMSCD is capable of reconstructing these scenes with mean accuracies of approximately 20–500% better than all competing algorithms adopted in this work. Furthermore, the reconstruction efficiency of trace materials in the scene has been assessed: it is shown that the KMSCD is capable of recovering ~12% better than that of the C-SCD. These results suggest that the proposed DL using a simple clustering method for the construction of the dictionary has been shown to enhance the scene reconstruction substantially. When the proposed KMSCD is incorporated with the Fast non-negative orthogonal matching pursuit (FNNOMP) to constrain the maximum number of materials to coexist in a pixel to four, experiments have shown that it achieves approximately ten times better than that constrained by using the widely employed TMM algorithm. This may suggest that the proposed DL method using KMSCD and together with the FNNOMP will be more suitable to be the material allocation module of HSI scene simulators like the CameoSim package.

Method for time-of-flight estimation of low frequency acoustic signals in reverberant and noisy environment with sparse impulse response

“…In order to statistically evaluate the various sparse deconvolution algorithms from section 3 for TOF estimation application, 300 pseudorandom signals are generated for each algorithm and for each SNR, and then TOF estimation is performed according to section 4. In the following, the Greedy algorithms NNOMP [41,42], FNNOMP [43] and sNMF [26] of the family of OMP as well as NNOLS and sNNOLS [41,42] of the family of OLS and the convex approach NLARS [44] are tested. In particular, the influence of the k reconstructed peaks in h is investigated.…”

Using a Panchromatic Image to Improve Hyperspectral Unmixing

“…Nevertheless, these NMF methods are highly sensitive to their initialization, especially when applied to an entire image subject to spectral variability [16,35,36]. New approaches have also been proposed, like in the K-Means Sparse Coding Dictionary (KMSCD) algorithm [37], which relies on a K-means clustering [38] to initialize a Sparse Coding Dictionary method [39]. These approaches exhibit the benefit of a clustering step in hyperspectral unmixing.…”

Hyperspectral and Multispectral Image Fusion by Deep Neural Network in a Self-Supervised Manner

“…They treat the fusion of LR HSI and HR MSI as an ill-posed inverse problem. Equations are first established by observation model [21][22][23] and then constrained by many handcraft priors. Popular priors contain the sparse prior [16,18,19] and lowrankness prior [24,25].…”

“…These equations are solved by iterative optimization methods such as alternating direction method of multipliers (ADMM) [26] and gradient descent algorithm [27]. Dictionary learning methods [22,23] are a representative kind of VM-based methods. By using sparse representation, they can combine the dictionaries from LR HSI and the high-resolution sparse coefficients from HR MSI to obtain HR HSI.…”