Search

Precise and timely classification of land cover types plays an important role in land resources planning and management. In this paper, nine kinds of land cover types in the acquired hyperspectral scene are classified based on the kernel collaborative representation method. To reduce the spectral shift caused by adjacency effect when mining the spatial-spectral features, a correlation coefficient-weighted spatial filtering operation is proposed in this paper. Additionally, by introducing this operation into the kernel collaborative representation method with Tikhonov regularization (KCRT) and discriminative KCRT (DKCRT) method, respectively, the weighted spatial-spectral KCRT (WSSKCRT) and weighted spatial-spectral DKCRT (WSSDKCRT) methods are constructed for land cover classification. Furthermore, aiming at the problem of difficulty of pixel labeling in hyperspectral images, this paper attempts to establish an effective land cover classification model in the case of small-size labeled samples. The proposed WSSKCRT and WSSDKCRT methods are compared with four methods, i.e., KCRT, DKCRT, KCRT with composite kernel (KCRT-CK), and joint DKCRT (JDKCRT). The experimental results show that the proposed WSSKCRT method achieves the best classification performance, and WSSKCRT and WSSDKCRT outperform KCRT-CK and JDKCRT, respectively, obtaining the OA over 94% with only 540 labeled training samples, which indicates that the proposed weighted spatial filtering operation can effectively alleviate the spectral shift caused by adjacency effect, and it can effectively classify land cover types under the situation of small-size labeled samples.

Computational intelligence is a very active and fruitful research of artificial intelligence with a broad spectrum of applications. Remote sensing data has been a salient field of application of computational intelligence algorithms, both for the exploitation of the data and for the research/ development of new data analysis tools. In this editorial paper we provide the setting of the special issue “Computational Intelligence in Remote Sensing” and an overview of the published papers. The 11 accepted and published papers cover a wide spectrum of applications and computational tools that we try to summarize and put in perspective in this editorial paper.

Multispectral (MS) and hyperspectral (HS) images have been successfully and widely used in remote sensing applications such as target detection, change detection, and anomaly detection. In this paper, we aim at reviewing recent change detection papers and raising some challenges and opportunities in the field from a practitioner&rsquo, s viewpoint using MS and HS images. For example, can we perform change detection using synthetic hyperspectral images? Can we use temporally-fused images to perform change detection? Some of these areas are ongoing and will require more research attention in the coming years. Moreover, in order to understand the context of our paper, some recent and representative algorithms in change detection using MS and HS images are included, and their advantages and disadvantages will be highlighted.

The current increase of spatial as well as spectral resolutions of modern remote sensing sensors represents a real opportunity for many prac tical applications but also generates important challenges in terms of image processing. In particular, the spatial correlation between pixels and/or the spectral correlation between spectral bands of a given pixel cannot be ignored. The traditional pixel-based representation of images does not facilitate the handling of these correlations. In this paper, we discuss the inter est of a particular hierarchical region-based representation of images based on binary partition tree (BPT). This representation approach is very flexible as it can be applied to any type of image. Here both optical and radar images will be discussed. Moreover, once the image representation is computed, it can be used for many different applications. Filtering, segmentation, and classifica- tion will be detailed in this paper. In all cases, the interest of the BPT representation over the classical pixel-based representa- tion will be highlighted

[EN] Nowadays, image analysis is becoming more important because of its ability to perform fast and non-invasive low-cost analysis on products and processes. Image analysis is a wide denomination that encloses classical studies on gray scale or RGB images, analysis of images collected using few spectral channels (sometimes called multispectral images) or, most recently, data treatments to deal with hyperspectral images, where the spectral direction is exploited in its full extension. Pioneering data treatments in image analysis were applied to simple images mainly for defect detection, segmentation and classification by the Computer Science community. From the late 80s, the chemometric community joined this field introducing powerful tools for image analysis, which were already in use for the study of classical spectroscopic data sets and were appropriately modified to fit the particular characteristics of image structures. These chemometric approaches adapt to images of all kinds, from the simplest to the hyperspectral images, and have provided new insights on the spatial and spectroscopic information of this kind of data sets. New fields open by the introduction of chemometrics on image analysis are exploratory image analysis, multivariate statistical process control (monitoring), multivariate image regression or image resolution. This paper reviews the different techniques developed in image analysis and shows the evolution in the information provided by the different methodologies, which has been heavily pushed by the increasing complexity of the image measurements in the spatial and, particularly, in the spectral direction. © 2011 Elsevier B.V., We would like to thank the Chemolab Editorial Board for inviting us to prepare this paper. This research was supported by the Spanish Government (Science and Innovation Ministry) under projects DPI2008-06880-C03-03 and CTQ2009-11572.

This paper studies a semi-supervised Bayesian unmixing algorithm for hyperspectral images. This algorithm is based on the normal compositional model recently introduced by Eismann and Stein. The normal compositional model assumes that each pixel of the image is modeled as a linear combination of an unknown number of pure materials, called endmembers. However, contrary to the classical linear mixing model, these endmembers are supposed to be random in order to model uncertainties regarding their knowledge. This paper proposes to estimate the mixture coefficients of the Normal Compositional Model (referred to as abundances) as well as their number using a reversible jump Bayesian algorithm. The performance of the proposed methodology is evaluated thanks to simulations conducted on synthetic and real AVIRIS images.

Hyperspectral image (HSI) classification is gaining a lot of momentum in present time because of high inherent spectral information within the images. However, these images suffer from the problem of curse of dimensionality and usually require a large number samples for tasks such as classification, especially in supervised setting. Recently, to effectively train the deep learning models with minimal labelled samples, the unlabeled samples are also being leveraged in self-supervised and semi-supervised setting. In this work, we leverage the idea of semi-supervised learning to assist the discriminative self-supervised pretraining of the models. The proposed method takes different augmented views of the unlabeled samples as input and assigns them the same pseudo-label corresponding to the labelled sample from the downstream task. We train our model on two HSI datasets, namely Houston dataset (from data fusion contest, 2013) and Pavia university dataset, and show that the proposed approach performs better than self-supervised approach and supervised training., Comment: The paper was submitted in IGARSS, 2023 conference and is not accepted to appear in the proceedings. The page requirement is 4 pages, including references

This paper proposes to use band selection-based dimensionality reduction (BS-DR) technique in addressing a challenging multi-temporal hyperspectral images change detection (HSI-CD) problem. The aim of this work is to analyze and evaluate in detail the CD performance by selecting the most informative band subset from the original high-dimensional data space. In particular, for cases where ground reference data are available or unavailable, either supervised or unsupervised CD approaches are designed. The following sub-problems in HSI-CD are investigated, including: (1) the estimated number of multi-class changes; (2) the binary CD; (3) the multiple CD; (4) the estimated optimal number of selected bands; and (5) computational efficiency. The main contribution of this paper is to provide for the first time a thorough analysis of the impacts of band selection on the HSI-CD problem, thus to fix the gap in the state-of-the-art techniques either by simply utilizing the full dimensionality of the data or exploring a complex hierarchical change analysis. It is applicable to CD problems in multispectral or PolSAR images when the feature space is expanded for discriminant feature extraction. Two real multi-temporal hyperspectral Hyperion datasets are used to validate the proposed approaches. Quantitative and qualitative experimental results demonstrated that by selecting a subset of the most informative and distinct spectral bands, the proposed approaches offered better CD performance than the state-of-the-art techniques using original full bands, without losing the change representative and discriminable capabilities of a detector.

Biodiversity monitoring is an almost inconceivable challenge at the scale of the entire Earth. The current (and soon to be flown) generation of spaceborne and airborne optical sensors (i.e., imaging spectrometers) can collect detailed information at unprecedented spatial, temporal, and spectral resolutions. These new data streams are preceded by a revolution in modeling and analytics that can utilize the richness of these datasets to measure a wide range of plant traits, community composition, and ecosystem functions. At the heart of this framework for monitoring plant biodiversity is the idea of remotely identifying species by making use of the ‘spectral species’ concept. In theory, the spectral species concept can be defined as a species characterized by a unique spectral signature and thus remotely detectable within pixel units of a spectral image. In reality, depending on spatial resolution, pixels may contain several species which renders species-specific assignment of spectral information more challenging. The aim of this paper is to review the spectral species concept and relate it to underlying ecological principles, while also discussing the complexities, challenges and opportunities to apply this concept given current and future scientific advances in remote sensing.

Object tracking using Hyperspectral Images (HSIs) obtains satisfactory result in distinguishing objects with similar colors. Yet, the tracking algorithm tends to fail when the target undergoes deformation. In this paper, a SiamRPN based hyperspectral tracker is proposed to deal with this problem. Firstly, a band selection method based on a genetic optimization method is designed for rapidly reducing the redundancy of information in HSIs. Specifically, three bands with highest joint entropy are selected. To solve the problem that the information of the template in the SiamRPN model decays over time, an update network is trained on the dataset from general objective tracking benchmark, which can obtain effective cumulative templates. The use of cumulative templates with spectral information makes it easier to track the deformed target. In addition, transfer learning of the pre-trained SiamRPN is designed to obtain a better model for HSIs. The experimental results show that the proposed tracker can obtain good tracking results over the entire public dataset, and that it is better than the other popular trackers when the target’s deformation is qualitatively and quantitatively compared, achieving an overall success rate of 57.5% and a deformation challenge success rate of 70.8%.

The low spatial resolution of hyperspectral images means that existing mixed pixels rely heavily on spectral information, making it difficult to differentiate between the target of interest and the background. The endmember extraction method is powerful in enhancing spatial structure information for hyperspectral anomaly detection, whereas most approaches are based on matrix representation, which inevitably destroys the spatial or spectral information. In this paper, we treated the hyperspectral image as a third-order tensor and proposed a novel anomaly detection method based on a low-rank linear mixing model of the scene background. The obtained abundance maps possessed more distinctive features than the raw data, which was beneficial for identifying anomalies in the background. Specifically, the low-rank tensor background was approximated as the mode-3 product of an abundance tensor and endmember matrix. Due to the distinctive features of the background’s abundance maps, we characterized them by tensor regularization and imposed low rankness through CP decomposition, smoothness, and sparsity. In addition, we utilized the ℓ1,1,2-norm to characterize the tube-wise sparsity of the anomaly, since it accounted for a small portion of the scene. The experimental results obtained using five real datasets demonstrated the outstanding performance of our proposed method.

Hyperspectral unmixing has attracted considerable attentions in recent years and some promising algorithms have been developed. In this paper, collaborative representation–based unmixing (CRU) for hyperspectral images is proposed. Different from imposing the sparseness constraint on training samples in sparse representation, collaborative representation emphasizes the collaboration of training samples. Furthermore, its closed form solution greatly improves computational efficiency. In the experiments, synthetic and the real hyperspectral data are used to evaluate the effectiveness and efficiency of the proposed collaborative representation-based hyperspectral unmixing algorithm.

Hyperspectral Images are a challenging area because of big dimensions, spatial and spectral low resolution. This paper aims to propose a method to provide an accurate and quick classification based on tensor decomposition and map segmentation. For reducing the dimensions of hyperspectral data and obtain the abundances map we used Parafac decomposition. For determining the segmentation abundances map representing the segmentation abundances matrix, an algorithm based on the Dynamic Clustering Algorithm [1] principle was used having as input the estimated abundances matrix. The segmentation abundances map was visually labeled using the true color image.

Terrain classification is an important research direction in the field of remote sensing. Hyperspectral remote sensing image data contain a large amount of rich ground object information. However, such data have the characteristics of high spatial dimensions of features, strong data correlation, high data redundancy, and long operation time, which lead to difficulty in image data classification. A data dimensionality reduction algorithm can transform the data into low-dimensional data with strong features and then classify the dimensionally reduced data. However, most classification methods cannot effectively extract dimensionality-reduced data features. In this paper, different dimensionality reduction and machine learning supervised classification algorithms are explored to determine a suitable combination method of dimensionality reduction and classification for hyperspectral images. Soft and hard classification methods are adopted to achieve the classification of pixels according to diversity. The results show that the data after dimensionality reduction retain the data features with high overall feature correlation, and the data dimension is drastically reduced. The dimensionality reduction method of unified manifold approximation and projection and the classification method of support vector machine achieve the best terrain classification with 99.57% classification accuracy. High-precision fitting of neural networks for soft classification of hyperspectral images with a model fitting correlation coefficient (R2) of up to 0.979 solves the problem of mixed pixel decomposition.

An accurate identification of objects from the acquisition system depends on the clear segmentation and classification of remote sensing images. With the limited financial resources and the high intra-class variations, the earlier proposed algorithms failed to handle the sub-optimal dataset. The building of an efficient training set iteratively in active learning (AL) approaches improves classification performance. The heuristics-based AL provides better results with the inheritance of contextual information and the robustness to noise variations. The uncertainty exists pixel variations make the heuristics-based AL fail to handle the remote sensing image classification. Previously, we focused on the extraction of clear textural pattern information by using the extended differential pattern-based relevance vector machine (EDP-AL). This paper extends that work into the novel pixel-certainty activity learning (PCAL) based on the information about textural patterns obtained from the extended differential pattern (EDP). Initially, distributed intensity filtering (DIF) is used to eliminate noise from the image, and then histogram equalization (HE) is used to improve the image quality. The EDP is used to merge and classify different labels for each image sample, and this algorithm expresses the textural information. The PCAL technique is used to classify the HSI patterns that are important in remote sensing applications using this pattern collection. Pavia University and Indian Pines (IP) are the datasets used to validate the performance of the proposed PCAL (PU). The ability of PCAL to accurately categorize land cover types is demonstrated by a comparison of the proposed PCAL with existing algorithms in terms of classification accuracy and the Kappa coefficient.

Due to sensor instability and atmospheric interference, hyperspectral images (HSIs) often suffer from different kinds of noise which degrade the performance of downstream tasks. Therefore, HSI denoising has become an essential part of HSI preprocessing. Traditional methods tend to tackle one specific type of noise and remove it iteratively, resulting in drawbacks including inefficiency when dealing with mixed noise. Most recently, deep neural network-based models, especially generative adversarial networks, have demonstrated promising performance in generic image denoising. However, in contrast to generic RGB images, HSIs often possess abundant spectral information; thus, it is non-trivial to design a denoising network to effectively explore both spatial and spectral characteristics simultaneously. To address the above issues, in this paper, we propose an end-to-end HSI denoising model via adversarial learning. More specifically, to capture the subtle noise distribution from both spatial and spectral dimensions, we designed a Residual Spatial-Spectral Module (RSSM) and embed it in an UNet-like structure as the generator to obtain clean images. To distinguish the real image from the generated one, we designed a discriminator based on the Multiscale Feature Fusion Module (MFFM) to further improve the quality of the denoising results. The generator was trained with joint loss functions, including reconstruction loss, structural loss and adversarial loss. Moreover, considering the lack of publicly available training data for the HSI denoising task, we collected an additional benchmark dataset denoted as the Shandong Feicheng Denoising (SFD) dataset. We evaluated five types of mixed noise across several datasets in comparative experiments, and comprehensive experimental results on both simulated and real data demonstrate that the proposed model achieves competitive results against state-of-the-art methods. For ablation studies, we investigated the structure of the generator as well as the training process with joint losses and different amounts of training data, further validating the rationality and effectiveness of the proposed method.

The high spatial dimensionality of the remote sensing images that are captured by modern hyperspectral sensors prevents many algorithms from being computationally feasible. Superpixel segmentation is a process that groups pixels into connected regions that are uniform according to one or more similarity measures. WaterPixel (WP) segmentation is a particular case of superpixel segmentation based on the watershed transform. In this paper an efficient implementation of the WP algorithm for the segmentation of remote sensing hyperspectral images on multi-core CPUs and programmable GPUs is explored. The proposed approach focuses on reducing the cost of the morphological gradient and the watershed segmen-tation, which are the two most costly steps of the algorithm., This work was supported in part by the Civil Program UAVs Initiative, promoted by the Xunta de Galicia and developed in partnership with the Babcock Company to promote the use of unmanned technologies in civil services. We also have to acknowledge the support by Ministerio de Ciencia e Innovación, Government of Spain (grant number PID2019-104834GB-I00), and Consellería de Educación, Universidade e Formación Profesional (ED431C 2018/19, and accreditation 2019-2022 ED431G-2019/04). All are cofunded by the European Regional Development Fund (ERDF).

Dimension reduction (DR) is an essential preprocessing for hyperspectral image (HSI) classification. Recently, nonnegative matrix factorization (NMF) has been shown as an effective tool for the DR of hyperspectral data given the fact that it provides interpretable results. However, the basic NMF ignores the geometric structure information of the HSI data, thus limiting its performance. To this end, a novel regularized NMF method, termed NMF with adaptive graph regularizer (NMFAGR), is proposed for the spectral-spatial dimension reduction of hyperspectral data in this paper. Specifically, to enhance the preservation ability of the geometric structure information, the NMFAGR performs the dimension reduction and graph learning simultaneously. Regarding the mutual correlation between these two tasks, a graph regularizer is added as an interaction. Moreover, to effectively utilize complementary information among spectral-spatial features, the NMFAGR allocates feature weight factors automatically without requiring any additional parameters. An efficient algorithm is utilized to solve the optimization problem. The effectiveness of the proposed method is demonstrated on three benchmark hyperspectral data sets through experimentation.

In this paper, a graph-based nonlocal total variation method (NLTV) is proposed for unsupervised classification of hyperspectral images (HSI). The variational problem is solved by the primal-dual hybrid gradient (PDHG) algorithm. By squaring the labeling function and using a stable simplex clustering routine, an unsupervised clustering method with random initialization can be implemented. The effectiveness of this proposed algorithm is illustrated on both synthetic and real-world HSI, and numerical results show that the proposed algorithm outperforms other standard unsupervised clustering methods such as spherical K-means, nonnegative matrix factorization (NMF), and the graph-based Merriman-Bence-Osher (MBO) scheme.

Crops, like other plants, clearly react to various changes in both natural and anthropogenic factors (herbicides, pesticides, fertilizers, etc.), which affects the amount of phytomass, its fractional composition, and developmental and physiological state of the plant, and, accordingly, is reflected in the spectral image. Data on spectral characteristics of plants allow users to determine quickly and with a high degree of reliability various indicators of the state of agricultural crops and thus improve the efficiency of agrotechnical practices and the use of land resources and facilitate the implementation of the precision farming concept. Reflective properties of plants (and hence crops) carry a large amount of meaningful information about the species, stage of development, and morpho-physiological state, allowing determination of the interrelations between the spectrometric characteristics and temporal physiological parameters. The paper presents the results of monitoring of the state of winter wheat and corn in experimental fields in southern and central Russia in the spring and summer of 2016.

International audience; For one given scene, multimodal data are acquired from multiple sensors. They share some similarities across the sensor types (redundant part of the information, also called coupling part) and they also provide modality-specific information (dissimilarities across the sensors, also called decoupling part). Additional critical knowledge about the scene can hence be extracted, which is not extractable from each modality alone. For the processing of multimodal data, we propose in this paper a model to simultaneously learn the underlying low-dimensional manifold in each modality, and locally align these manifolds across different modalities. For each pair of modalities we first build a common manifold that represents the corresponding (redundant) part of information, ignoring non-corresponding (modality specific) parts. We propose a semi-supervised learning model, using a limited amount of prior knowledge about the coupling and decoupling components of the different modalities. We propose a localized version of Laplacian eigenmaps technique specifically designed to handle multimodal manifold learning, in which the ideas of local patching of the manifolds, also known as manifold charting, is combined with the joint spectral analysis of the graph Laplacians of the different modalities. The limited given supervised information is then extending on the manifold of each modality. The idea of functional mapping is finally used to align the different manifolds across modalities. The evaluation of the proposed model using synthetic and real-world multimodal problems shows promising results, compared to several related techniques.

As the volume of remotely sensed data grows significantly, content-based image retrieval (CBIR) becomes increasingly important, especially for cloud computing platforms that facilitate processing and storing big data in a parallel and distributed way. This paper proposes a novel parallel CBIR system for hyperspectral image (HSI) repository on cloud computing platforms under the guide of unmixed spectral information, i.e., endmembers and their associated fractional abundances, to retrieve hyperspectral scenes. However, existing unmixing methods would suffer extremely high computational burden when extracting meta-data from large-scale HSI data. To address this limitation, we implement a distributed and parallel unmixing method that operates on cloud computing platforms in parallel for accelerating the unmixing processing flow. In addition, we implement a global standard distributed HSI repository equipped with a large spectral library in a software-as-a-service mode, providing users with HSI storage, management, and retrieval services through web interfaces. Furthermore, the parallel implementation of unmixing processing is incorporated into the CBIR system to establish the parallel unmixing-based content retrieval system. The performance of our proposed parallel CBIR system was verified in terms of both unmixing efficiency and accuracy.

Hyperspectral imaging techniques are becoming one of the most important tools to remotely acquire fine spectral information on different objects. However, hyperspectral images (HSIs) require dedicated processing for most applications. Therefore, several machine learning techniques were proposed in the last decades. Among the proposed machine learning techniques, unsupervised learning techniques have become popular as they do not need any prior knowledge. Specifically, sparse subspace-based clustering algorithms have drawn special attention to cluster the HSI into meaningful groups since such algorithms are able to handle high dimensional and highly mixed data, as is the case in real-world applications. Nonetheless, sparse subspace-based clustering algorithms usually tend to demand high computational power and can be time-consuming. In addition, the number of clusters is usually predefined. In this paper, we propose a new hierarchical sparse subspace-based clustering algorithm (HESSC), which handles the aforementioned problems in a robust and fast manner and estimates the number of clusters automatically. In the experiment, HESSC is applied to three real drill-core samples and one well-known rural benchmark (i.e., Trento) HSI datasets. In order to evaluate the performance of HESSC, the performance of the new proposed algorithm is quantitatively and qualitatively compared to the state-of-the-art sparse subspace-based algorithms. In addition, in order to have a comparison with conventional clustering algorithms, HESSC’s performance is compared with K-means and FCM. The obtained clustering results demonstrate that HESSC performs well when clustering HSIs compared to the other applied clustering algorithms.

Artificial Neural Networks (ANNs) have been used in a wide range of applications for complex datasets with their flexible mathematical architecture. The flexibility is favored by the introduction of a higher number of connections and variables, in general. However, over-parameterization of the ANN equations and the existence of redundant input variables usually result in poor test performance. This paper proposes a superstructure-based mixed-integer nonlinear programming method for optimal structural design including neuron number selection, pruning, and input selection for multilayer perceptron (MLP) ANNs. In addition, this method uses statistical measures such as the parameter covariance matrix in order to increase the test performance while permitting reduced training performance. The suggested approach was implemented on two public hyperspectral datasets (with 10% and 50% sampling ratios), namely Indian Pines and Pavia University, for the classification problem. The test results revealed promising performances compared to the standard fully connected neural networks in terms of the estimated overall and individual class accuracies. With the application of the proposed superstructural optimization, fully connected networks were pruned by over 60% in terms of the total number of connections, resulting in an increase of 4% for the 10% sampling ratio and a 1% decrease for the 50% sampling ratio. Moreover, over 20% of the spectral bands in the Indian Pines data and 30% in the Pavia University data were found statistically insignificant, and they were thus removed from the MLP networks. As a result, the proposed method was found effective in optimizing the architectural design with high generalization capabilities, particularly for fewer numbers of samples. The analysis of the eliminated spectral bands revealed that the proposed algorithm mostly removed the bands adjacent to the pre-eliminated noisy bands and highly correlated bands carrying similar information.

This paper introduces a feature selection method based on random forest -the Guided Regularized Random Forest (GRRF)- which can be used in classification and regression tasks. The method is based on the regularization of the information gain in the random forest nodes to obtain a subset of relevant and non-redundant features. The proposed method is used as a preliminary step In the process of retrieving biophysical parameters from a hyperspectral image. Preliminary experiments show that we can reduce the RMSE of the retrievals by around 7% for the Leaf Area Index and around 8% for the fraction of vegetation cover when compared to the results using random forest features.

This paper aimed to present a new method for the spectral-spatial classification of hyperspectral images, based on the idea of modified minimum spanning forest (MMSF). MMSF works on the obtained regions of pre-segmentation step that are considered as nodes of an image graph. In the proposed method, the image is first smoothed by the multiscale edge-preserving filter (MSEPF) and then the MMSF is built in each scale. Finally, all the classification maps of each scale are combined with a majority vote rule. The suggested method, named as MSEPF-MMSF, is performed on four hyperspectral images with different properties, and the experiments deal with the impacts of parameters of filter and the number of markers. The results demonstrate that the proposed method has improved the classification accuracies with respect to the previous methods.

The large volume of hyperspectral images (HSI) generated creates huge challenges for transmission and storage, making data compression more and more important. Compressive Sensing (CS) is an effective data compression technology that shows that when a signal is sparse in some basis, only a small number of measurements are needed for exact signal recovery. Distributed CS (DCS) takes advantage of both intra- and inter- signal correlations to reduce the number of measurements needed for multichannel-signal recovery. HSI can be observed by the DCS framework to reduce the volume of data significantly. The traditional method for estimating endmembers (spectral information) first recovers the images from the compressive HSI and then estimates endmembers via the recovered images. The recovery step takes considerable time and introduces errors into the estimation step. In this paper, we propose a novel method, by designing a type of coherent measurement matrix, to estimate endmembers directly from the compressively observed HSI data via convex geometry (CG) approaches without recovering the images. Numerical simulations show that the proposed method outperforms the traditional method with better estimation speed and better (or comparable) accuracy in both noisy and noiseless cases.

Numerous environmental and financial benefits of using hyperspectral imaging have driven much increased applications on plant monitoring and diagnosis. This paper is concerned with analysis of hyperspectral images for plant discrimination by means of their spectral and texture properties. The main contribution of the work lies in the use feature selection and Markov random field model (MRF) to facilitate such spectral-texture analysis to enhance prediction performance, as compared to conventional analysis methods. A hyperspectral dataset on control and stressed Arabidopsis plant leaves captured by a proximal hyperspectral imaging system was used in the experiment. Texture parameters with different orders were estimated from the MRF model and two support vector machine settings were used in the evaluation. Experimental results showed significant improvements in classification performance of the proposed spectral-texture approach over the conventional analysis methods.

Spectral unmixing extracts subpixel information by decomposing observed pixel spectra into a collection of constituent spectra signatures and their associated fractions. Considering the restriction of linear unmixing model, nonlinear unmixing algorithms find their applications in complex scenes. Kernel-based algorithms serve as important candidates for nonlinear unmixing as they do not require specific model assumption and have moderate computational complexity. In this paper we focus on the linear mixture and nonlinear fluctuation model. We propose a two-step kernel-based unmixing algorithm to address the case where a large spectral library is used as the candidate endmembers or the sparse mixture case. The sparsity-inducing regularization is introduced to perform the endmember selection and the candidate library is then pruned to provide more accurate results. Experimental results with synthetic and real data, particularly those laboratory-created labeled, show the effectiveness of the proposed algorithm compared with state-of-art methods.

International audience; Hyperspectral imaging, which records a detailed spectrum of light for each pixel, provides an invaluable source of information regarding the physical nature of the different materials, leading to the potential of a more accurate classification. However, high dimensionality of hyperspectral data, usually coupled with limited reference data available, limits the performances of supervised classification techniques. The commonly used pixel-wise classification lacks information about spatial structures of the image. In order to increase classification performances, integration of spatial information into the classification process is needed. In this paper, we propose to extend the watershed segmentation algorithm for hyperspectral images, in order to define information about spatial structures. In particular, several approaches to compute a one-band gradient function from hyperspectral images are proposed and investigated. The accuracy of the watershed algorithms is demonstrated by the further incorporation of the segmentation maps into a classifier. A new spectral-spatial classification scheme for hyperspectral images is proposed, based on the pixel-wise Support Vector Machines classification, followed by majority voting within the watershed regions. Experimental segmentation and classification results are presented on two hyperspectral images. It is shown in experiments that when the number of spectral bands increases, the feature extraction and the use of multidimensional gradients appear to be preferable to the use of vectorial gradients. The integration of the spatial information from the watershed segmentation in the hyperspectral image classifier improves the classification accuracies and provides classification maps with more homogeneous regions, compared to pixel-wise classification and previously proposed spectral-spatial classification techniques. The developed method is especially suitable for classifying images with large spatial structures.

Multispectral and Hyperspectral Images are wide-spread in Remote Sensing and Space Imaging Applications: they are typically three dimensional arrays of images and their raw content may include thousands of bands for a total of over a billion pixels. For adequate storing and transmission over satellite communication links a proper compression scheme is required. In this paper a FPGA implementation of the Consultative Committee for Space Data Systems (CCSDS) 123.0-B-1 standard algorithm is presented. The implementation results refer to a space-class device (Xilinx Virtex 5QV FX130T). The proposed architecture, composed by the Virtex5 FPGA and an external DRAM, implements circuital-oriented architectural improvements over the standard algorithm formulation. The proposed circuit, that is also user configurable in various aspects, overcomes the state of the art in terms of image pixel throughput (55.4 MSamples/s) and has a negligible footprint in terms of resource allocation (4%) for the selected FPGA.

Spectral unmixing aims at decomposing each image element of a hyperspectral scene in signals typically related to pure materials. This paper gives an added value to the results of this process by proposing Unmixing-based Denoising (UBD), a supervised methodology to recover bands characterized by a low Signal-to-Noise Ratio in a hyperspectral scene. In the first step of UBD, an unmixing procedure is carried out using a set of reference spectra which are noise-free, as they are averaged over areas for which ground truth is available. Results are inferred into the pixelwise reconstruction of a given band, expressed as a linear combination of the values of the reference spectra in that band, ignoring the residual vector which is mainly characterized by undesired atmospheric influences and sensor-induced noise. The reconstructed images exhibit both high visual quality and reduced spectral distortions.

Hyperspectral images are widely used in several real-life applications. In this paper, we investigate on the compression of hyperspectral images by considering different aspects, including the optimization of the computational complexity in order to allow implementations on limited hardware (i.e., hyperspectral sensors, etc.). We present an approach that relies on a three-dimensional predictive structure. Our predictive structure, 3D-MBLP, uses one or more previous bands as references to exploit the redundancies among the third dimension. The achieved results are comparable, and often better, with respect to the other state-of-art lossless compression techniques for hyperspectral images.

International audience; In this paper we present a feasibility study for a novel Raman microscopy set-up. Using polarized light for Raman analysis of the studied sample, along with spatial, rotational and spectral diversities, we aim at separating the contributions of the different micro-crystals present on the sample and estimate their spatial orientations. The approach is based on a three-way Candecomp/Parafac (CP) decomposition of the acquired data. Numerical simulations are used to validate the proposed method.

International audience; Linear unmixing of hyperspectral images is a popular approach to determine and quantify materials in sensed images. The linear unmixing problem is challenging because the abundances of materials to estimate have to satisfy non-negativity and full-additivity constraints. In this paper, we investigate an iterative algorithm that integrates these two requirements into the coefficient update process. The constraints are satisfied at each iteration without using any extra operations such as projections. Moreover, the mean transient behavior of the weights is analyzed analytically, which has never been seen for other algorithms in hyperspectral image unmixing. Simulation results illustrate the effectiveness of the proposed algorithm and the accuracy of the model.

In this paper we analyze the problem of tree species classification in the Southern Alps by using high geometrical resolution airborne hyperspectral data. In addition, we study the effects of downscaling the spectral resolution through the use of very high geometrical resolution satellite images. The analysis is carried out on data acquired over a mountain area in the Southern Alps. This area is characterized by eleven tree species, both coniferous and broadleaved, distributed in topographically complex site. For each data source a specific processing chain was developed and a Support Vector Machine classifier was used. The experimental results made it clear that airborne hyperspectral data are effective for tree species classification in complex mountain areas (kappa accuracy of about 0.78). The spectral downscaling to very high resolution satellite multispectral images allows one to keep the spatial detail of the analysis but reducing significantly the level of accuracy in class discrimination (acceptable results were obtained only for macro-classes of species, for which the kappa accuracy was 0.70).

The present paper addresses the problem of the classification of hyperspectral images with multiple imbalanced classes and very high dimensionality. Class imbalance is handled by resampling the data set, whereas PCA and a supervised filter are applied to reduce the number of spectral bands. This is a preliminary study that pursues to investigate the benefits of combining several techniques to tackle the imbalance and the high dimensionality problems, and also to evaluate the order of application that leads to the best classification performance. Experimental results demonstrate the significance of using together these two preprocessing tools to improve the performance of hyperspectral imagery classification. Although it seems that the most effective order corresponds to first a resampling strategy and then a feature (or extraction) selection algorithm, this is a question that still needs a much more thorough investigation in the future This work has partially been supported by the Spanish Ministry of Education and Science under grants CSD2007–00018, AYA2008–05965–0596 and TIN2009–14205, the Fundació Caixa Castelló–Bancaixa under grant P1–1B2009–04, and the Generalitat Valenciana under grant PROMETEO/2010/028

ISBN : 978-1-4244-2002-5; International audience; Segmenting dynamic contrast enhanced-MRI series of small animal, which are intrinsically noisy and low contrasted images with low resolution, is the aim of this paper. To do this, a segmentation method taking into account the temporal (spectral) and spatial information is presented on several series. The idea is to start from a temporal classification, and to build a probability density function of contours conditionally to this classification. Then, this function is segmented to find potentially tumorous areas. The method is presented on several series after a range normalization histogram in order to compare the series.

RESUMEN: En este artículo se presenta un nuevo método de procesado de imágenes hiperespectrales capturadas mediante un sensor espectroscópico de imagen. Basado en un algoritmo de clasificación lineal, permite mejorar las restricciones de tiempo real que planteaba un sistema de clasificación de materia prima diseñado previamente. El procesado de datos consta de dos bloques: compresión de características e interpretación de las mismas. La compresión se realiza empleando Análisis de Componentes Principales (PCA), mientras que la clasificación se realiza mediante el Mapeo del Ángulo Espectral (SAM). La técnica diseñada se ha aplicado al control de la calidad de la materia prima de la industria tabacalera. En ésta se pretende separar la materia prima (las hojas de tabaco) de un conjunto de materiales diversos como madera, plásticos, cartón, papel de caramelo, etc. El sensor espectroscópico empleado para la captura de las imágenes hiperespectrales consta de una cámara monocromática y un dispositivo basado en un elemento dispersivo pasivo Prisma – Red de difracción – Prisma. ABSTRACT: A data processing method for hyperspectral images from an imaging spectroscopic sensor is presented. It is a linear algorithm which comes to solve real time constrains of a previously designed raw material classification system. The new data processing method is composed of two blocks: data compression and classification unit. Data compression is performed by means of Principal Component Analysis (PCA) and the interpretation algorithm for classification is the Spectral Angle Mapper (SAM). This strategy has been successfully tested on the raw material on-line characterization in tobacco industry. In this application, the desired raw material (tobacco leaves) should be discriminated from other unwanted spurious materials, such as wood, plastic, cardboard, leather, candy paper, etc. Hyperspectral images are recorded by a spectroscopic sensor consisting of a monochromatic camera and a dispersive passive Prism-Grating-Prism device.

A data processing method to classify hyperspectral images from an imaging spectroscopic sensor is evaluated. Each image contains the whole diffuse reflectance spectra of the analyzed material for all the spatial positions along a specific line of vision. The implemented linear algorithm comes to solve real time constrains typical of industrial systems. This processing method is composed of two blocks: data compression is performed by means of principal component analysis (PCA) and the spectral interpretation algorithm for classification is the spectral angle mapper (SAM). This strategy, applying PCA and SAM, has been successfully tested for online raw material sorting in the tobacco industry, where the desired raw material (tobacco leaves) should be discriminated from other unwanted spurious materials, such as plastic, cardboard, leather, feathers, candy paper, etc. Hyperspectral images are recorded by a sensor consisting of a monochromatic camera and a passive prism-grating-prism device. Performance results are compared with a spectral interpretation algorithm based on artificial neural networks (ANN).

This paper investigates on the development of an advanced method for lossless compression of hyperspectral data to be implemented on board of a space platform. An adaptive Differential Pulse Code Modulation (DPCM) method, jointly exploiting spectral and spatial correlation and utilizing space-oriented context-based entropy coding, is taken as starting point. The algorithm considered utilizes a "classified" DPCM approach, in which predictors, taking into account the statistical properties of the data being compressed, are preliminarily calculated and then adaptively selected or combined. Two fuzzy clustering algorithms are tested with the aim of finding the best algorithm to be employed in the initialization phase, which is the core of the "classified" DPCM compression procedure, may be performed off line so as to unaffect the computational complexity of the online procedure running on board. The final method utilizes a standard CCSDS-Rice space encoder and represents a good tradeoff between compression capability and computational complexity. Overall coding performances, as well as differences between the two fuzzy clustering algorithms, are reported and discussed through extensive experiments carried out on four hyperspectral AVIRIS images.