Recently, the selective structured state space model (S6) built upon the Mamba architecture has attracted widespread attention for its outstanding performance in long-range modeling. However, existing Mamba-based hyperspectral image (HSI) classification methods suffer from certain limitations in extracting edge information. To address this problem, a new HSI classification framework called Pinwheel and Wavelet-based Spatial-Spectral Mamba (S 2 PW-Mamba) is proposed in this paper. The input to our S 2 PW-Mamba is the complete, unsegmented image, which preserves the correlation between pixels in the HSI while leveraging convolution to enhance the extraction of local information. Specifically, two S6 modules are employed along different scanning directions to extract the spatial-spectral features from the HSI, acting on the spectral and spatial domains, respectively. Meanwhile, a GateFusion module is designed to adaptively guide the fusion of spatial and spectral features by discerning their relative importance. More precisely, our S 2 PW-Mamba first extracts both edge and central spatial features using a Pinwheel-shaped convolution and a four-directional spatial scanning module, effectively capturing spatial-contextual relationships. Subsequently, by integrating the wavelet transform with the S6 architecture, the model is capable of capturing both global contextual dependencies and local texture details within the spectral domain. Comprehensive experiments on three benchmark HSI datasets show that our approach outperforms existing state-of-the-art methods.

The distinguishability between background and anomaly is significant for accurate hyperspectral anomaly detection (HAD). The property that background and anomaly are characterized as signals with distinct differences in the frequency domain is of great value, however, existing HAD algorithms rarely consider this. In addition, deep learning (DL)-based methods that exploit reconstruction errors for HAD can inadvertently reconstruct anomalies alongside the background, leading to a high false alarm rate (FAR). To address these challenges, this study proposes a novel HAD model based on frequency-spatial–spectral domain analysis and criss-cross transformer (CCTrans), named F$\mathbf {S}^{\mathbf {2}}$CCTrans. Specifically, to suppress anomaly reconstruction, the frequency domain analysis paradigm is integrated into HAD, an advanced saliency map (SM) is constructed by analyzing the discriminative characteristics of anomaly and background in both amplitude and phase spectrum. The SM guides the model training in the direction of suppressing its capability to express and reconstruct anomalies, thereby obtaining a pure background estimate. To craft a superb background generator, the CCTrans network is designed that captures spatial-spectral features of the background in a very effective and efficient way. The CCTrans incorporates an ingenious criss-cross attention mechanism, which is focused on aggregating contextual information of all the pixels along its criss-cross path, thus significantly decreasing computational burden. Extensive comparisons with eight state-of-the-art methods on synthetic and real datasets demonstrate the superiority of F$\mathbf {S}^{\mathbf {2}}$CCTrans. Meanwhile, ablation studies reveal that the CCTrans requires about 68% less GPU memory and about 58% fewer FLOPs, highlighting its efficiency.

Multispectral and hyperspectral image fusion has emerged as a highly effective technique for obtaining images with both high spatial and spectral resolution. This is an ill-posed problem that poses significant challenges to the optimization solution, which is often mitigated by incorporating low-rank and smooth priors to restrict the solution space. Traditionally, these priors are combined additively using nuclear norm and total variation (TV) regularization. However, the intricate interactions between these priors make it difficult to accurately characterize the prior structure using an additive approach. Moreover, their influence is heavily dependent on the trade-off parameter between the regularization terms. To address this issue, we propose a novel fusion method leveraging a joint low-rank and smooth tensor prior (LRST). The LRST method introduces a tensor nuclear norm on the gradient maps of various dimensions of the target image capitalizing on the analogous manifold structures shared between the original image and its gradient map, which seamlessly integrates the two priors into a unified regularization framework. This facilitates the precise and convenient exploitation of spatial and spectral correlations inherent in the desired hyperspectral image. Experimental findings demonstrate that compared to state-of-the-art fusion methods, the LRST approach yields finely fused images.

Normalized Difference Vegetation Index (NDVI) data from optical satellites have been widely used in the field of remote sensing. However, due to cloud cover and other extreme weather reasons, NDVI time series are often missing. In order to fill the observation gap under severe weather conditions, an NDVI time series reconstruction method that fuses optical and synthetic aperture radar (SAR) images has been developed. However, optical images often suffer from long-term data missing, severely limiting the usability of this reconstruction method. To solve this problem, this paper proposes a NDVI-like image strategy to avoid reliance on optical data. Specifically, the strategy obtains NDVI-like images by fitting the multivariate linear relationship between VV, VH polarization bands of SAR images and NDVI images. Then, the optical image information is represented by NDVI-like images, and the SAR and NDVI-like time series are input into the time series reconstruction network to reconstruct the NDVI time series in cloudy areas. Experimentation has revealed that the NDVI-like image strategy can effectively supplements optical image information for the NDVI reconstruction network, reduce the the influence of SAR image speckle noise on reconstruction outcomes, and obtain better reconstruction effects.

Medium spatial resolution satellite image series is important for some applications. Currently, the most widely used medium resolution data is from Sentinel-2 and Landsat-8 satellites. For the purpose of acquiring dense image series, the fusion of them has drawn much attention. In this field, the deep learning-based fusion models have appeared excellent performance, which demonstrated the applicability for deep learning for the fusion. However, existing deep learning based fusion models all need high quality and abundant training data, which may limit their applications. To address this problem, we develop an unsupervised fusion model based on convolutional neural networks (CNNs), which can fuse the Sentinel-2 and Landsat images without other training data. This fusion model takes full advantage of CNNs in the image fusion task, but get raid of the limitation of the requirement of training data. In the experiment, we test the new model using the Sentinel-Landsat dataset, and compare it with other two fusion approaches. The experimental results indicate that the proposed fusion model has remarkable fusion performance especially in terms of accuracy. However, it is can not capture the surface changes well, which is the main direction for us to improve it in the future.

Graph-based methods have excellent performance in hyperspectral image (HSI) classification because of their strong ability to explore the relationship between labeled and unlabeled samples. However, most graph-based methods do not take sufficient account of label information and more discriminant features to establish graph connections, which will lead to a lot of improper connections, and then produce over-smooth or noisy classification results. To solve this issue, we propose a posterior probability fused superpixel graph (PPSPG), which exploits the fitting ability of supervised models to encode information from labels and discriminant features into the posterior probabilities. By measuring the weights between the posterior probabilities of any two superpixels, the proposed PPSPG can fully integrate label information into graph connetions, and alleviate the over-smoothing between superpixels. Experiments on real hyperspectral data show that our method has outstanding accuracies and can obtain quite distinct classification boundaries.

In the field of hyperspectral unmixing, deep learning has received increasing attention due to its powerful learning and data representation capabilities. Autoencoder is a popular technique for unmixing. Recently, an autoencoder-based depth image prior algorithm has been proposed for hyperspectral unmixing, which employs geometric methods to extract endmembers. The performance of this network solely focuses on estimating the abundance of images, and it has achieved good unmixing results. However, the depth image only employs one core network, which makes it highly vulnerable to noise and can lead to unstable unmixing outcomes. To address the aforementioned issue, this paper proposes a collaborative consistency autoencoder-based hyperspectral unmixing approach with deep image prior (CCAUDIP). For the proposed CCAUDIP model, it adopts two autoencoders to cooperatively handle the same input data to enhance the generalization ability of the network. Additionally, a consistency constraint is introduced to restrict the abundance outputs of the two core autoencoders and improve the robustness of the network. The experimental results show that the CCAUDIP method can achieve better unmixing results compared to other advanced unmixing algorithms.

Spectral unmixing is a crucial aspect of hyperspectral image processing. Given the low spatial resolution of hyperspectral remote sensing sensors, combined with the complexity and diversity of actual ground objects, hyperspectral remote sensing images often contain numerous mixed pixels, which make spectral unmixing a challenging task. Recent advancements in spectral libraries have shown promising results for decomposing mixed pixels in hyperspectral remote sensing images. Sparse unmixing, a semi-supervised unmixing strategy, avoids the drawbacks of blind source unmixing algorithms, which may extract virtual endmembers with no physical meaning. In this paper, we propose the multiscale spatial sparse unmixing (MSSU) algorithm, which utilizes the signal adaptive spatial multiscale unmixing of the over-segmentation method to decompose the complex unmixing problem. Furthermore, weighting factors are introduced to extract spatial information from the spectral image. The experimental results obtained from simulated hyperspectral datasets reveal the great potential of the proposed algorithm in unmixing.

As the spectral library continues to expand, sparse hyperspectral unmixing methods have been developed to solve the mixing problem without the need for end-member extraction or generation. These methods leverage the intrinsic spectral and spatial information to enhance the accuracy of fractional abundance estimation. However, their effectiveness is limited by rigid spatial regularization and insufficient utilization of spectral spatial information, which hampers the improvement of unmixing performance. To overcome this limitation, we present a novel algorithm named Local Spectral Similarity Guided Sparse Hyperspectral Unmixing with Spatial Graph Regularization (SGSU). In SGSU, we introduce a spatial graph regularization to enforce the inter-pixel correlation within spatial clusters and assign them to corresponding abundance vectors. To reduce the computational cost, we employ an adaptive superpixel-based spatial grouping strategy to segment the hyperspectral image, which translates the intrinsic geometry into constraints on abundance. Furthermore, we introduce a weighting factor with two components into the sparse unmixing framework. One component is based on the row sparsity of the estimated abundances, indicating the presence of active end-members; the other component is based on the similarity between neighboring pixels, which promotes piecewise smoothness of the estimated abundances. To obtain a more robust solution, we adopt a double-loop scheme based on the alternating direction method of multipliers (ADMM) algorithm to solve the SGSU model. Experimental results on both simulated and real hyperspectral datasets demonstrate that the proposed SGSU algorithm outperforms state-of-the-art sparse unmixing methods in terms of both accuracy of abundance estimation and end-member identification from spectral libraries. Our algorithm achieves superior unmixing results, which indicates its potential for practical applications in hyperspectral imaging.

Convolutional neural networks (CNNs) have been widely employed for hyperspectral image (HSI) classification due to their powerful ability to extract local spatial features. However, CNN-based methods cannot establish long-range dependencies among sequences of pixels. Transformers offer significant advantages when processing sequential data and can establish global relationships, but they still encounter a number of challenges, such as their limited spatial feature extraction ability, or their high computational cost. In order to address the aforementioned issues, we develop a new fast HSI classification approach combining transformers and SimAM-based CNNs. The latter are utilized to extract better spatial features, where the complex spatial characteristics of HSIs are retrieved using an improved hierarchical 2D dense network structure. A dual attention unit (DAU) mechanism is then utilized to direct the model’s attention to discriminative spatial pixel characteristics and effective feature map channels, while suppressing information that is irrelevant for classification purposes. Regarding the spectral features, after extracting hierarchical local characteristics from various convolutional layers (using the hierarchical dense network structure), a squeezed-enhanced axial transformer is employed to establish global long-range dependencies whilst enhancing the ability of the model to extract local detail features in the HSI. Besides, a new Lion optimizer is utilized to improve the classification performance of our model. Our quantitative and comparative experiments on four benchmark datasets demonstrate the effectiveness of the proposed approach provides better classification results than other state-of-the-art approaches. Moreover, our FTSCN also achieves better classification results than other methods in practical scenarios.

Sparse unmixing is a semi-supervised method whose pur-pose is to find the best subset of library entries from the spec-tral library that best model the image. In sparse unmixing, the current main development direction is to incorporate the spatial information of the image into the model. Existing spa-tial sparse unmixing algorithms mainly use spatial weights or spatial regularization to characterize the spatial correlation between pixels to improve the unmixing results. For the complex and diverse hyperspectral data in reality, most al-gorithms are only good at processing a single scene, which brings greater challenges to their practicality. In order to ad-dress this issue, a new dual spatial weighted sparse unmixing model (DSWSU) is proposed, which simultaneously ex-ploits the spatially homogeneous information of images. For the proposed DSWSU, a pre-calculated superpixel weighting factor is designed to mitigate the effect of noise on unmixing. Meanwhile, the spatial neighborhood weighting factor aims to promote the local smoothness of the abundance maps. As a simple unmixing model, the proposed DSWSU can be quickly solved by the alternating direction multiplier method (ADMM). Experimental results on simulated hyperspectral data indicate that the proposed DSWSU method can achieve accurate abundance estimation in various scenarios (low or high noise interference), and obtain better unmixing results than other state-of-the-art unmixing algorithms.

Spectral unmixing is an essential technology for the interpretation of hyperspectral remote sensing images. Sparse unmixing has become a research hotspot in the field of spectral unmixing since it circumvents the issue of endmember extraction. Regularization based on spatial information further improves the performance of sparse unmixing. However, multiple regularization terms increase the complexity of the sparse model and the difficulty of regularization parameter tuning. To overcome this drawback, a new dual reweighted low-rank and total variation sparse unmixing (DRLRSU-TV) method is proposed, which jointly imposes the low-rank constraint, dual reweighted sparse constraint and total variation (TV) regularizer on the classic sparse unmixing model via two regularization terms. Experiment results on simulated hyper-spectral data verify the excellent unmixing performance of the proposed algorithm compared to other state-of-the-art sparse unmixing methods.

Since traditional convolutional neural network (CNN) is used to extract the spectral-spatial features of hyperspectral image (HSI) will result in lots of spatial information redundancy. Octave convolution is used to replace the traditional CNN to reduce spatial redundancy and expand the receptive field. However, the methods based on 3D octave convolution may cause many parameters and the model to be complicated. To address these issues, we propose an HSI classification approach based on a double-branch multi-scale spectral-spatial convolution network (DBMS) in this paper. Firstly, We utilize 2D octave convolution and 3D DenseNet sub-networks with different kernels sizes to extract complex spatial features and spectral features, respectively. Furthermore, a channel attention module and a spectral attention module are employed in this two sub-network respectively, to highlight the important feature areas and specific spectral bands that consist of significant information for the classification. Compared with several other state-of-the-art methods, our proposed method can achieve competitive performance on Salinas Valley (SV) HSI dataset.

In the field of hyperspectral unmixing (HU), deep learning (DL) techniques have attracted increasing attention due to their powerful capabilities in learning and feature extraction. The autoencoder framework has flexible scalability as well as good unsupervised learning ability, and it achieves good performance in hyperspectral unmixing. However, some traditional autoencoder-based algorithms only focus on pixel-level reconstruction loss, which ignores the detailed information contained in the material. In addition, these algorithms usually only use a single autoencoder with non-convex properties, which brings great difficulty to the solution. In this paper, a cascaded autoencoders-based spectral-spatial unmixing (CASSU) framework is proposed to address these issues. For the proposed CASSU, on the one hand, two concatenated autoencoders are introduced to better find the global optimal solution. On the other hand, in this cascaded deep network, spectral angle mapping (SAM) and convolution operations are used to extract the spectral-spatial information of the image. Experimental results on real hyperspectral data indicate that the newly proposed CASSU algorithm has better unmixing performance compared to several state-of-the-art unmixing algorithms.

Hyperspectral unmixing is an important image interpre-tation technique that aims to estimate the pure constituent materials (endmembers) and their corresponding fractional abundances in each mixed pixel. Nonnegative matrix factorization (NMF) has attracted a lot of attention because of its ability to solve mixed pixel scenarios. The sparse NMF method achieves better unmixing results thanks to its full use of the sparse characteristics of the data. However, most existing sparse NMF unmixing techniques lack the consid-eration of spatial information. In fact, hyperspectral images contain intrinsic geometric information as well as rich spatial information. In this paper, a spatial graph regularized nonneg-ative matrix factorization unmixing framework (SGNMF) is established. For the proposed SGNMF, on the one hand, the graph regularization is introduced to characterize the latent manifold structure of the data, and on the other hand, the spatial weighting factor is used to mine the spatial correlation between pixels. The optimization problem of the SGNMF model can be solved by a multiplicative iterative rule. Exper-imental results on synthetic data sets indicate that the newly proposed SGNMF method is able to produce better results than other advanced spectral unmixing algorithms.

Nowadays, a variety of different sources can be combined together to measure and monitor maritime human activities. Reliable data fusion techniques are essential to associate the targets from different systems for maritime surveillance. In particular, the fusion of data from Sentinel-2 satellites and the Automatic Identification System (AIS) has attracted wide attention due to their public availability and complementarity. However, most traditional methods for target association are not suitable for this particular case, due to the time lag phenomenon of Sentinel-2 data. In this study, we first construct two new datasets for the detection of moving ships and their wakes based on Sentinel-2 images. Combined with the detection results obtained by the You Only Look Once (YOLOv5) model, the position and course information of the detected ships are first extracted. After carefully analyzing the time lag phenomenon of Sentinel-2 data, we develop a new domain adaptation-based method for target association based on the fusion of Sentinel-2 and AIS data, called Moving Ship Optimal Association (MSOA). Different from standard domain adaptation methods only for representation alignment, the proposed MSOA is able to align representation, time and position simultaneously. A case study is provided in which the newly proposed method is tested over the Port of Long Beach, USA. Experimental results demonstrate that both moving ships and wakes are well detected. Specifically, our newly proposed MSOA exhibits more accurate and robust performance when compared to traditional methods, and the detected ships without corresponding AIS tracks can also be detected by our MSOA. Moreover, the real sensing time and time lag of Sentinel-2 data are deduced with high accuracy. Overall, it can be concluded that our MSOA provides a new perspective for accurate target association based on heterogeneous data fusion.

Remotely sensed hyperspectral images contain several bands (at about adjoining frequencies) for a similar zone on the surface of the Earth. Hyperspectral unmixing is a significant method for breaking down hyperspectral images into the components (endmembers) that conform each (potentially mixed) pixel and their abundance maps. Nonnegative matrix factorization (NMF) has attracted huge consideration because of the way that it can address mixed pixel scenarios. Most existing NMF unmixing techniques do not include spatial information in the analysis. An ongoing trend is to fuse the spatial and the spectral information contained in hyperspectral scenes to improve the solution. In this article, we build up another hyperspectral unmixing technique named spectral–spatial weighted sparse NMF (SSWNMF), in which two weighting factors are acquainted into the NMF model to upgrade the sparsity of the solution and capture the piecewise smooth structure of the data. We adopt a multiplicative iterative strategy to implement the proposed SSWNMF model. Our experimental results, conducted with both synthetic and real hyperspectral data, uncover that the proposed SSWNMF strategy can get accurate unmixing results over those gave by other unmixing strategies, with less parameter tuning.

Spatio-temporal fusion has been developed to generate the synthetic remote sensing data with high spatial resolution and high temporal resolution simultaneously. To date, a number of spatio-temporal fusion methods have been developed and most of them are remarkable. However, most the methods are designed to achieve better fusion performance and higher fusion accuracy, while the fusion speed is always ignored. As a matter of fact, most current spatio-temporal fusion methods are time-consuming. To address this defect of spatiotemporal fusion, in this paper we propose a extremely fast spatiotemporal fusion method. The core idea of this method is extracting the spatial information from the prior high-spatial-resolution images and embedding that into the low-spatial-resolution images by local normalization to predict the missing high-spatial-resolution images. In the experiments, two dataset, including a Landsat-MODIS dataset and a Sentienl-MODIS dataset, are adopted to testing this method. The experimental results demonstrate this method can achieve great performance with extremely fast speed.

With the increase of available spectral libraries, sparse unmixing has attracted great attention in the field of hyperspectral image unmixing. When the spatial information is integrated into the traditional sparse unmixing model, it achieves better performance. However, the less accurate description of the spatial structure limits the performance of the previous spatial sparse unmixing methods. To address this limitation, a new technique called superpixel based low-rank sparse unmixing (SpLRSU) is established, which encourages the local spatial consistency and the spatial continuity of the image. Specifically, superpixel segmentation is used to adaptively generate local homogeneous regions, and then the low-rank constraint is enforced on the abundance vectors of each spatial group to preserve the low-dimensional structure of superpixel blocks. Meanwhile, the spectral-spatial weighted sparse regularization term is introduced to promote the sparsity of fractional abundances in the spectral and spatial domains. The experimental results on the synthetic data set show that the newly proposed algorithm is superior to other advanced sparse unmixing algorithms.

Spectral unmixing is an important technique for hyperspectral image application, which aims to estimate the pure spectral signatures in each mixed pixel and their corresponding fractional abundances. However, due to the influence of factors such as illumination, topography change and atmosphere, spectral variability is inevitable, which will lead to inaccurate unmixing results. Traditional unmixing methods fail to handle this problem, especially the complex spectral variability in the image. To address this limitation, a new technique called low-rank subspace unmixing (LRSU) was established, which aims to jointly estimate a subspace projection and abundance maps. For the proposed LRSU approach, the original data is projected into a low-rank subspace to deal with various spectral variabilities in spectral unmixing. Meanwhile, the spectral-spatial weighted sparse regularization term is introduced to upgrade the sparsity of the solution and capture the piecewise smooth structure of the data. The experimental results, conducted using synthetic data sets, quantitatively indicate that the proposed LRSU strategy produces better results than other advanced spectral unmixing methods.

Hyperspectral unmixing is a technique for selecting endmembers (pure spectral constituents) and their abundances (proportions). Recently, sparse unmixing is a semi-supervised method in which mixed pixels are represented in the form of combinations of a number of pure spectral signatures from a large spectral library. Compared with other methods, the sparse unmixing method exhibits significant advantages. However, most of these sparse unmixing methods were implemented in spatial domain, where the information is too scattered, redundant and susceptible to noise. In this paper, we propose a new unmixing method called spectral-spatial weighted sparse unmixing in the transform domain (SSTSU) to impose the abundance sparsity and enhance the anti-noise performance. The experimental results show that the proposed algorithm has better anti-noise performance and unmixing results compared with other advanced sparse unmixing methods.

Hyperspectral unmixing aims to decompose a hyperspectral image (HSI) into a collection of constituent materials, or end-members, and their corresponding abundance fractions. Recently, nonnegative tensor factorization (NTF)-based spectral unmixing methods have attracted significant attention owing to their outstanding performance when representing an HSI without any information loss. However, tensor factorization-based HSI methods do not fully exploit the spatial contextual information present in the scene. Besides, these approaches are sensitive to low signal-to-noise ratio (SNR) in HSIs. To address this limitation, we propose a new spectral-spatial weighted sparse nonnegative tensor factorization (SSWNTF) method to preserve the spatial details in the abundance maps via the spectral and spatial weighting factors. Our experiments with simulated data sets certified that the proposed method outperforms other advanced methods.

Sparse representation-based approaches have been successfully applied to remotely sensed hyperspectral image unmixing. In recent years, sparse unmixing techniques have incorporated spatial information into the sparse unmixing model, achieving improved fractional abundance results. Most spatial-based sparse unmixing methods utilize regular-shaped neighborhoods (e.g., a cross or a square window) to characterize the spatial-contextual information around each pixel. However, the spatial characteristics of natural scenes are not always uniform, but vary according to the observed objects. Therefore, assuming uniform spatial neighborhoods may not be consistent with real spatial structures in the scene. Super-pixels offer a good solution to this problem since they can better characterize such spatial structures. Based on this observation, in this paper we develop a new superpixel-guided sparse unmixing (SPGSU) method for hyperspectral scenes. The proposed SPGSU includes the spatial correlation through a superpixel-based technique rather than assuming predefined pixel grids. Each superpixel can be regarded as a small spatial region, whose shape and size can be adaptively changed to accommodate different spatial structures. Our experimental results, conducted using simulated data sets, quantitatively indicate that our newly proposed method produces better results than other advanced spectral unmixing methods.

Spectral unmixing aims at estimating the fractional abundances of a set of pure spectral materials (endmembers) in each pixel of a hyperspectral image. The wide availability of large spectral libraries has fostered the role of sparse regression techniques in the task of characterizing mixed pixels in remotely sensed hyperspectral images. A general solution for sparse unmixing methods consists of using the l2regularizer to control the sparsity, resulting in a very promising performance but also suffering from sensitivity to large and small sparse coefficients. A recent trend to address this issue is to introduce weighting factors to penalize the nonzero coefficients in the unmixing solution. While most methods for this purpose focus on analyzing the hyperspectral data by considering the pixels as independent entities, it is known that there exists a strong spatial correlation among features in hyperspectral images. This information can be naturally exploited in order to improve the representation of pixels in the scene. In order to take advantage of the spatial information for hyperspectral unmixing, in this paper, we develop a new spectral-spatial weighted sparse unmixing (S2WSU) framework, which uses both spectral and spatial weighting factors, further imposing sparsity on the solution. Our experimental results, conducted using both simulated and real hyperspectral data sets, illustrate the good potential of the proposed S2WSU, which can greatly improve the abundance estimation results when compared with other advanced spectral unmixing methods.

Spectral unmixing is an important technique for remotely sensed hyperspectral image interpretation, of which the goal is to decompose the image into a set of pure spectral components (endmembers) and their abundance fractions in each pixel of the scene. Sparse-representation-based approaches have been widely studied for remotely sensed hyperspectral unmixing. A recent trend is to incorporate the spatial information to improve the spectral unmixing results. Those methods generally assume that the abundances of the pixels are piecewise smooth and fall into a homogeneous region occupied by the same endmembers and their corresponding fractional abundances. However, in real scenarios, abundances may vary abruptly from pixel to pixel. Therefore, the former assumption in most spatial models does not hold. To address this limitation, we propose a new strategy to preserve the spatial details in the abundance maps via a spatial discontinuity weight. Our experimental results, conducted with both simulated and real hyperspectral data sets, illustrate the good potential of our discontinuity-preserving strategy for sparse unmixing, which can greatly improve the abundance estimation results.

Sparse unmixing of hyperspectral data is an important technique which aims at estimating the fractional abundances of endmembers (pure spectral components). It is well known that enforcing sparseness becomes a necessary process in sparse unmixing methods. To better exploit the sparsity in hyperspectral imagery, a double reweighted sparse unmixing algorithm has been proposed. However, it focusses on analyzing the hyperspectral data without fully incorporating the spatial information. To address this limitation, a spatial weighted sparse unmixing (SWSU) algorithm is proposed in this paper, which can take full advantage of the spatial information and further enhance the sparsity of the abundances. This is done by incorporating local neighborhood weights into the double reweighted sparse unmixing formulation. Experimental results on simulated hyperspectral data sets illustrate the good potential of the spatial weighted strategy for sparse unmixing introduced in this paper, which can greatly improve abundance estimation results.

Morphological attribute profiles (MAPs) are one of the most effective methodologies to characterize the spatial information in remote sensing images. This technique extracts components able to accurately describe objects in the surface of the Earth. In this work, we present a new method for impervious surface extraction from multispectral images using morphological attribute profiles. The proposed method first uses morphological profiles to extend Landsat ETM+ images with additional features. Then, we adopt a vegetation-impervious surface-soil (V-I-S) model and extract three pure classes (endmembers) from these images (i.e. vegetation, impervious surface and soil) using the vertex component algorithm (VCA). Finally, linear spectral mixture analysis (SMA) is conducted to extract the impervious surface percentage (ISP). To test the performance of the proposed method, more than 300 test samples including business districts, residential areas and urban roads are randomly selected from QuickBird imagery with very high resolution. The coefficient of determination R2is 0.7571, which significantly outperformed other standard techniques in the literature. The obtained experimental results demonstrate that the proposed approach based on morphological attribute profiles can lead to very good extraction and characterization of impervious surfaces.

Most blind hyperspectral unmixing methods exploit convex geometry properties of hyperspectral data. The minimum volume simplex analysis (MVSA) is one of such methods, which, as many others, estimates the minimum volume (MV) simplex where the measured vectors live. MVSA was conceived to circumvent the matrix factorization step often implemented by MV-based algorithms and also to cope with outliers, which compromise the results produced by MV algorithms. Inspired by the recently proposed robust MV enclosing simplex (RMVES) algorithm, we herein introduce the robust MVSA (RMVSA), which is a version of MVSA robust to noise. As in RMVES, the robustness is achieved by employing chance constraints, which control the volume of the resulting simplex. RMVSA differs, however, substantially from RMVES in the way optimization is carried out. In this paper, we develop a linearization relaxation of the nonlinear chance constraints, which can greatly lighten the computational complex of chance constraint problems. The effectiveness of RMVSA is illustrated by comparing its performance with the state of the art.

In online service systems, the delay experienced by users from service request to service completion is one of the most critical performance metrics. To improve user delay experience, recent industrial practices suggest a modern system design mechanism: proactive serving, where the service system predicts future user requests and allocates its capacity to serve these upcoming requests proactively. This approach complements the conventional mechanism of capability boosting. In this paper, we propose queuing models for online service systems with proactive serving capability and characterize the user delay reduction by proactive serving. In particular, we show that proactive serving decreases average delay exponentially (as a function of the prediction window size) in the cases where service time follows light-tailed distributions. Furthermore, the exponential decrease in user delay is robust against prediction errors (in terms of miss detection and false alarm) and user demand fluctuation. Compared with the conventional mechanism of capability boosting, proactive serving is more effective in decreasing delay when the system is in the light-load regime. Our trace-driven evaluations demonstrate the practical power of proactive serving: for example, for the data trace of light-tailed YouTube videos, the average user delay decreases by 50% when the system predicts 60 s ahead. Our results provide, from a queuing-theoretical perspective, justifications for the practical application of proactive serving in online service systems.

Spectral unmixing is an important technique for hyperspectral data exploitation. In order to solve the unmixing problem using a collection of previously available spectral signatures (i.e., a spectral library), sparse unmixing aims at finding the optimal subset of endmembers to represent the pixels in a hyperspectral image. The classic collaborative unmixing globally assumes that all pixels in a hyperspectral scene share the same active set of endmembers. This assumption rarely holds in practice, as endmembers tend to appear localized in spatially homogeneous areas rather than spread over the whole image. To address this limitation, in this letter, we introduce a new strategy to preserve local collaborativity for sparse hyperspectral unmixing. The proposed approach, which is called local collaborative sparse unmixing, considers the fact that endmember signatures generally appear distributed in local spatial regions instead of uniformly distributed throughout the scene. The proposed approach, which includes spatial information in the standard collaborative formulation, has been experimentally validated using both simulated and real hyperspectral data sets.

Motivated by the increasing popularity of learning and predicting human user behavior in communication and computing systems, in this paper, we investigate the fundamental benefit of predictive scheduling, i.e., predicting and pre-serving arrivals, in controlled queueing systems. Based on a lookahead-window prediction model, we first establish a novel queue-equivalence between the predictive queueing system with a fully efficient scheduling scheme and an equivalent queueing system without prediction. This result allows us to analytically demonstrate that predictive scheduling necessarily improves system delay performance and drives it to zero with increasing prediction power. It also enables us to exactly determine the required prediction power for different systems and study its impact on tail delay. We then propose the Predictive Backpressure (PBP) algorithm for achieving optimal utility performance in such predictive systems. PBP efficiently incorporates prediction into stochastic system control and avoids the great complication due to the exponential state space growth in the prediction window size. We show that PBP achieves a utility performance that is within O(ε) of the optimal, for any ε > 0, while guaranteeing that the system delay distribution is a shifted-to-the-left version of that under the original Backpressure algorithm. Hence, the average delay under PBP is strictly better than that under Backpressure, and vanishes with increasing prediction window size. This implies that the resulting utility-delay tradeoff with predictive scheduling can beat the known optimal [O(ε),O(log(1/ε))] tradeoff for systems without prediction. We also develop the Predictable-Only PBP (POPBP) algorithm and show that it effectively reduces packet delay in systems where traffic can only be predicted but not pre-served.

Motivated by the increasing popularity of learning and predicting human user behavior in communication and computing systems, in this paper, we investigate the fundamental benefit of predictive scheduling, i.e., predicting and pre-serving arrivals, in controlled queueing systems. Based on a lookahead-window prediction model, we first establish a novel queue-equivalence between the predictive queueing system with a fully-efficient scheduling scheme and an equivalent queueing system without prediction. This result allows us to analytically demonstrate that predictive scheduling necessarily improves system delay performance and drives it to zero with increasing prediction power. It also enables us to exactly determine the required prediction power for different systems and study its impact on tail delay. We then propose the Predictive, Backpressure, (PBP) algorithm for achieving optimal utility performance in such predictive systems. PBP efficiently incorporates prediction into stochastic system control and avoids the great complication due to the exponential state space growth in the prediction window size. We show that PBP achieves a utility performance that is within O(ε) of the optimal, for any ε>0, while guaranteeing that the system delay distribution is a shifted-to-the-left version of that under the original Backpressure algorithm. Hence, the average delay under PBP is strictly better than that under Backpressure, and vanishes with increasing prediction window size. This implies that the resulting utility-delay tradeoff with predictive scheduling can beat the known optimal [O(ε), O(log(1/ε))] tradeoff for systems without prediction.

In online service systems, delay experienced by a user from the service request to the service completion is one of the most critical performance metrics. To improve user delay experience, in this paper, we investigate a novel aspect of system design: proactive serving, where the system can predict future user request arrivals and allocate its capacity to serve these upcoming requests proactively. In particular, we investigate the average user delay under proactive serving from a queuing theory perspective. We show that proactive serving reduces the average user delay exponentially (as a function of the prediction window size) under M/M/1 queueing models. Our simulation results show that, for G/G/1 queueing models, the average user delay also decreases significantly under proactive serving.

We study the problem of maximizing the broadcast rate in peer-to-peer (P2P) systems under node degree bounds, i.e., the number of neighbors a node can simultaneously connect to is upper-bounded. The problem is critical for supporting high-quality video streaming in P2P systems and is challenging due to its combinatorial nature. In this paper, we address this problem by providing the first distributed solution that achieves near-optimal broadcast rate under arbitrary node degree bounds and over arbitrary overlay graph. It runs on individual nodes and utilizes only the measurement from their one-hop neighbors, making the solution easy to implement and adaptable to peer churn and network dynamics. Our solution consists of two distributed algorithms proposed in this paper that can be of independent interests: a network-coding-based broadcasting algorithm that optimizes the broadcast rate given a topology, and a Markov-chain guided topology hopping algorithm that optimizes the topology. Our distributed broadcasting algorithm achieves the optimal broadcast rate over arbitrary P2P topology, while previously proposed distributed algorithms obtain optimality only for P2P complete graphs. We prove the optimality of our solution and its convergence to a neighborhood around the optimal equilibrium under noisy measurements or without time-scale separation assumptions. We demonstrate the effectiveness of our solution in simulations using uplink bandwidth statistics of Internet hosts.

Broadcasting systems such as P2P streaming systems represent important network applications that support up to millions of online users. An efficient broadcasting mechanism is at the core of the system design. Despite substantial efforts on developing efficient broadcasting algorithms, the following important question remains open: How to achieve the maximum broadcast rate in a distributed manner with each user maintaining information queues only for its direct neighbors? In this work, we first derive an innovative formulation of the problem over acyclic overlay networks with arbitrary underlay capacity constraints. Then, based on the formulation, we develop a distributed algorithm to achieve the maximum broadcast rate and every user only maintains one queue per-neighbor. Due to its lightweight nature, our algorithm scales very well with the network size and remains robust against high system dynamics. Finally, by conducting simulations we validate the optimality of our algorithm under different network capacity models. Simulation results further indicate that the convergence time of our algorithm grows linearly with the network size, which suggests an interesting direction for future investigation.

We study the problem of maximizing the broadcast rate in peer-to-peer (P2P) systems under node degree bounds, i.e., the number of neighbors a node can simultaneously connect to is upper-bounded. The problem is critical for supporting high-quality video streaming in P2P systems, and is challenging due to its combinatorial nature. In this paper, we address this problem by providing the first distributed solution that achieves near-optimal broadcast rate under arbitrary node degree bounds, and over arbitrary overlay graph. It runs on individual nodes and utilizes only the measurement from their one-hop neighbors, making the solution easy to implement and adaptable to peer churn and network dynamics. Our solution consists of two distributed algorithms proposed in this paper that can be of independent interests: a network-coding based broadcasting algorithm that optimizes the broadcast rate given a topology, and a Markov-chain guided topology hopping algorithm that optimizes the topology. Our distributed broadcasting algorithm achieves the optimal broadcast rate over arbitrary P2P topology, while previously proposed distributed algorithms obtain optimality only for P2P complete graphs. We prove the optimality of our solution and its convergence to a neighborhood around the optimal equilibrium under noisy measurements or without timescale separation assumptions. We demonstrate the effectiveness of our solution in simulations using uplink bandwidth statistics of Internet hosts.

Two questions on the theory of content distribution capacity are addressed in this paper: What is the worst user delay performance bound in a chunk-based P2P streaming systems under peer fanout degree constraint? Can we achieve both the minimum delay and the maximum streaming rate simultaneously? In the homogeneous user scenario, we propose a tree-based algorithm called Inverse Waterfilling, which schedules the chunk transmission following an optimal transmitting structure, under fanout degree bound. We show that the algorithm guarantees the delay bound for each chunk of the stream and maintains the maximum streaming rate at the same time.