New effective spectral matching measures for hyperspectral data analysis

The authors propose an information theoretic criterion, called spectral information divergence (SID) for spectral similarity and discriminability. It is derived from the concept of divergence arising in information theory and can be used to describe the statistics of a spectrum. Unlike spectral angle mapper (SAM) which extracts geometric features between two spectra, SID views each pixel spectrum as a random variable and then measures the discrepancy of probabilistic behaviors between two spectra. In order to evaluate SID, SAM is used for comparison via hyperspectral data. Experimental results show that SID can characterise spectral similarity and variability more effectively than SAM.

A hyperspectral image can be considered as an image cube where the third dimension is the spectral domain represented by hundreds of spectral wavelengths. As a result, a hyperspectral image pixel is actually a column vector with dimension equal to the number of spectral bands and contains valuable spectral information that can be used to account for pixel variability, similarity and discrimination. We present a new hyperspectral measure, the spectral information measure (SIM), to describe spectral variability and two criteria, spectral information divergence and spectral discriminatory probability for spectral similarity and discrimination, respectively. The spectral information measure is an information-theoretic measure which treats each pixel as a random variable using its spectral signature histogram as the desired probability distribution. Spectral information divergence (SID) compares the similarity between two pixels by measuring the probabilistic discrepancy between two corresponding spectral signatures. The spectral discriminatory probability calculates spectral probabilities of a spectral database (library) relative to a pixel to be identified so as to achieve material identification. In order to compare the discriminatory power of one spectral measure relative to another, a criterion is also introduced for performance evaluation, which is based on the power of discriminating one pixel from another relative to a reference pixel. The experimental results demonstrate that the new hyperspectral measure can characterize spectral variability more effectively than the commonly used spectral angle mapper (SAM).

The high spectral resolution of hyperspectral images (HSI) requires a heavy processing load. Assigning each pixel to a group in the image, which is called superpixel, and processing the superpixels instead of the pixels is resorted as a means to overcome this challenge in the hyperspectral literature. In this paper, we propose an algorithm to segment a hyperspectral image into superpixels by means of iteratively updating the boundary pixels of superpixels. We first explore the optimal similarity metric for the boundary pixel updates with the contraint of keeping the superpixel boundaries aligned with the object boundaries in the image. We investigate two approaches for similarity detection between pixels during this update, first comparing the hyperspectral pixels individually, and second, comparing the pixels by using also their neigborhood. The spectral similarity metrics used for investigation are selected as spectral angle mapping (SAM) [1], spectral information divergence (SID) [2] and spatial coherence distance [3] due to their common usage. The proposed approach is compared with a pioneer state-of-the-art superpixel algorithm, SLIC [4], and its superiority is verified in terms of the superpixelization performance metrics, namely boundary recall and undersegmentation error [5].

Object-based image analysis methods have been developed recently. They have since become a very active research topic in the remote sensing community. This is mainly because the researchers have begun to study the spatial structures within the data. In contrast, pixel-based methods only use the spectral content of data. To evaluate the applicability of object-based image analysis methods for land-cover information extraction from hyperspectral data, a comprehensive comparative analysis was performed. In this study, six supervised classification methods were selected from pixel-based category, including the maximum likelihood (ML), fisher linear likelihood (FLL), support vector machine (SVM), binary encoding (BE), spectral angle mapper (SAM) and spectral information divergence (SID). The classifiers were conducted on several features extracted from original spectral bands in order to avoid the problem of the Hughes phenomenon, and obtain a sufficient number of training samples. Three supervised and four unsupervised feature extraction methods were used. Pixel based classification was conducted in the first step of the proposed algorithm. The effective feature number (EFN) was then obtained. Image objects were thereafter created using the fractal net evolution approach (FNEA), the segmentation method implemented in eCognition software. Several experiments have been carried out to find the best segmentation parameters. The classification accuracy of these objects was compared with the accuracy of the pixel-based methods. In these experiments, the Pavia University Campus hyperspectral dataset was used. This dataset was collected by the ROSIS sensor over an urban area in Italy. The results reveal that when using any combination of feature extraction and classification methods, the performance of object-based methods was better than pixel-based ones. Furthermore the statistical analysis of results shows that on average, there is almost an 8 percent improvement in classification accuracy when we use the object-based methods.

Hyperspectral imaging is concerned with the measurement, analysis, and interpretation of spectra acquired from a given scene (or specific object) at a short, medium or long distance by an airbone or satellite sensor. Over the last few years, hyperspectral image data sets have been collected for a great amount of locations over the world, using a variety of instruments for Earth observation. Despite the increasing importance of hyperspectral images in remote sensing applications, there is no common repository of hyperspectral data intended to distribute and share hyperspectral data sets in the community. Quite opposite, the hyperspectral data sets which are available for public use are spread among different storage locations and present significant heterogeneity regarding the storage format, associated meta-data (if any), or ground-truth availability. As a result, the development of a standardized hyperspectral data repository is a highly desired goal in the remote sensing community. In this paper, we take a necessary first step towards the development of a digital repository for remotely sensed hyperspectral data. The proposed system allows uploading new hyperspectral data sets along with meta-data, ground-truth and analysis results, with the ultimate goal of sharing publicly available hyperspectral images within the remote sensing community. The database has been designed in order to allow storing relevant information for the hyperspectral data available through the system, including basic image characteristics (width, height, number of bands, format) and more advanced meta-data (ground-truth information, publications in which the data has been used). The current implementation consists of a front-end to ease the management of images through a web interface, thus containing both synthetic and real hyperspectral images from two highly representative instruments, such as NASAs Airborne Visible Infra-Red Imaging Spectrometer (AVIRIS) over the Cuprite Mining District in Nevada. Most importantly, the developed system includes a spectral unmixing-based content based image retrieval (CBIR) functionality which allows searching for images on the spectral unmixing information (spectrally pure components or endmembers and their associated abundances in the scene). This information is stored as meta-data associated to each hyperspectral image instance, and then used to search and retrieve images based on information content. This paper presents the design of the system and a preliminary validation of the unmixing-based retrieval functionality using both synthetic and real hyperspectral images stored in the database.

In volatile, conflict-affected regions, rapidly mapping asbestos-cement rooftops is critical to mitigate health risks from airborne fibres. Over a 4-month field campaign (Nov 2023–Apr 2024), we partnered with the Israel Space Agency and Ministry of Environmental Protection to acquire disturbance-free field spectra at multiple Kibbutzim, Moshavim and cities, using an ASD FieldSpec 4 High-Res with both the SoilPro® apparatus and contact probe to build a comprehensive spectral library (Sup Figs. 5–14). Leveraging EnMAP Level 2A hyperspectral imagery (17 May 2024), we applied MNF noise reduction, precise co-registration, and cloud/shadow masking before executing eight supervised classifiers; Linear Spectral Unmixing, Support Vector Machine, Spectral Angle Mapper, Adaptive Coherence Estimator, Mahalanobis Distance, Maximum Likelihood, Spectral Information Divergence, and Matched Filtering, in an iterative filtering cascade. Exhaustive ground-truth surveys across villages and cities achieved an 86% positive match rate despite urban complexity and security-driven coordinate restrictions. This integrative workflow combining rigorous field calibration, multi-algorithm spectral filtering, and exhaustive validation, demonstrates that orbit-based hyperspectral data can reliably map asbestos hazards at scale, guiding timely emergency response and long-term remediation in high-risk settings.

Integrated management of forest ecosystems requires an accurate and all-sided mastery of the forest information, of which forest ecosystem cover especially at tree species level is the most basic and important component. The study investigated and demonstrated the mapping potential of the forest ecosystem at tree species level from high spatial resolution hyperspectral images. The mapping performances of eight conventional classification methods including Maximum Likelihood (ML), Mahalanobis Distance (MaD), Minimum Distance (MD), Parallelepiped (P), Binary Encoding (BE), Spectral Angle Mapper (SAM), Spectral Information Divergence (SID) and Support Vector Machine (SVM), were verified based on two noise treatments (noise fraction and noise removal) and three leaf growth periods (tender leaf period, young leaf period and adult leaf period). It could be confirmed that noise removal obviously contributed to improving the classification agreement and young leaf period was most suitable for mapping forest ecosystem at tree species level from high spatial resolution hyperspectral images. ML, P, BE and SID were not considered appropriate according to good results with overall accuracy and kappa coefficient exceeding 85% and 0.80 respectively. Though MD also produced a very high classification agreement, it could not cover up its poor potential to identify tree species by spectral features. Even if SVML, SVMP, SVMR and SVMS performed the stablest and could generate good results across three periods, the best result was obtained by SAM. Except that the difference was significant between MD and SVMS at the 5% significance level in tender leaf period, the comparative tests did not provide more proof to show the significant difference between the methods considered appropriate.

This paper uses simulated annealing and focus on the spectral angle mapper (SAM), to demonstrate how the separability of two mean spectra from different species can be increased by choosing the bands that maximize the metric. It is known that classification performance is enhanced when the differences in mean spectra for each end-member species are maximized. Comparison was made using the selected bands derived from the proposed method, to all bands in the electromagnetic spectrum (EMS), only the bands in the visible, near infrared and short wave infrared regions of the EMS and selected bands using stepwise discriminant analysis. The bands from the proposed method often indicates a better choice of band selection as viewed by the summary statistics for (a) the SAM measurements, (b) the correlations between bands and (c) the spectral information divergence (SID), for each pair of species; and the classification accuracy of SAM and SID.

Different classification techniques are being designed and under usage for classification of hyper spectral images. The usage of these classifiers differ for different type of hyper spectral data and application domain. The performance of these classifiers are influenced by feature preprocessing stage. In this research work we have investigated the impact of feature preprocessing using Principal Component Analysis (PCA) and Linear Discriminant Analysis (LDA) on classification stage. The experiment is carried out using three sets of hyperspectral data. Classifications were carried out using three different classification techniques namely Maximum Likelihood Estimation (MLE), Constraint Energy Minimization (CEM) and Spectral Angle Mapper (SAM) on the preprocessed hyperspectral data. It is observed that the impact of PCA and LDA on the classification techniques are in two aspects (a) The preprocessing techniques facilitates to achieve high classification accuracy even with less number of training features and (b) Preprocessing expedites the classification process for large data sets. Also it can be concluded that PCA outperforms LDA in case of noisy data.

Abstract. In this paper a new classification technique for hyperspectral data based on synergetics theory is presented. Synergetics – originally introduced by the physicist H. Haken – is an interdisciplinary theory to find general rules for pattern formation through selforganization and has been successfully applied in fields ranging from biology to ecology, chemistry, cosmology, and thermodynamics up to sociology. Although this theory describes general rules for pattern formation it was linked also to pattern recognition. Pattern recognition algorithms based on synergetics theory have been applied to images in the spatial domain with limited success in the past, given their dependence on the rotation, shifting, and scaling of the images. These drawbacks can be discarded if such methods are applied to data acquired by a hyperspectral sensor in the spectral domain, as each single spectrum, related to an image element in the hyperspectral scene, can be analysed independently. The classification scheme based on synergetics introduces also methods for spatial regularization to get rid of "salt and pepper" classification results and for iterative parameter tuning to optimize class weights. The paper reports an experiment on a benchmark data set frequently used for method comparisons. This data set consists of a hyperspectral scene acquired by the Airborne Visible Infrared Imaging Spectrometer AVIRIS sensor of the Jet Propulsion Laboratory acquired over the Salinas Valley in CA, USA, with 15 vegetation classes. The results are compared to state-of-the-art methodologies like Support Vector Machines (SVM), Spectral Information Divergence (SID), Neural Networks, Logistic Regression, Factor Graphs or Spectral Angle Mapper (SAM). The outcomes are promising and often outperform state-of-the-art classification methodologies.

Hyperspectral techniques have made it possible the recognition and classification of tree species. However, it is unknown which bands are the most suitable to distinguish tree species. This study aimed at obtaining sensitive bands that can be used for tree species classification based on correlation analysis between the hyperspectral data and pigment contents of trees. In this study, a total of 264 spectral data sets were collected with a band range of 400 nm to 925 nm at fixed points and times form March 2010 to February 2011, including 171 and 93 pieces of branches and leaves for Cunninghamia Lanceolata and Pinus massoniana Lamb, respectively. The in situ hyperspectral reflectance data of these two tree species were collected at canopy using ASD (Analytical Spectral Devices) FieldSpec HandHeld spectrradiometer, a new product of ASD America Inc. At the same time, chlorophyll a&b, chlorophyll-a, chlorophyll-b, carotenoids and xanthophyll of branches and leaves for these species were measured in lab. First, the correlation between the hyperspectral data and the pigment contents was analyzed, and the spectral intervals of these two conifers with higher coefficients of correlation were obtained. The data of spectral intervals selected were used for classification of these tree species using five classification algorithms including Support Vector Machine-Radial Basis Function (SVM-RBF), BP neural network, Mahalanobis Distance, Bayes, and Spectral Angle Mapping (SAM). The results showed that the use of chlorophyll led to better performance of classification than other pigments. The sensitive spectral intervals for these two conifers based on chlorophyll ranged from 401 nm to 504 nm and from 659 nm to 686 nm. Among these five methods, it was found that SVM-RBF and BP neural network classifications resulted in better performance. Second, in order to test the dependence and adaptability for imaging spectral data, the obtained spectral intervals were combined using Gaussian curve instead of spectral response function (SRF) and the combined data was classified using the above five methods. The accuracy of classification obviously reduced around 15% to 20%. But, the performance of chlorophyll a&b could be still better than other pigments. Third, In order to test the applicability of the spectral intervals, we added 46 spectral data sets of Cinnamomum camphora, collected from 2004 to 2006, to do the classification. The performance of chlorophyll was also better than other pigments. This implied that the obtained spectral intervals 401 nm to 504 nm and 659 nm to 686 nm based on chlorophyll were consistent and could be applied to classification of other tree species. But, the spectral intervals were not narrow enough and there is a need of further study and examination.

Remote sensing detection of gold related alteration zones in Um Rus area, Central Eastern Desert of Egypt

We propose a new spectral-spatial method for hyperspectral image classification based on a graph cut. The classification task is formulated as an energy minimization problem on the graph of image pixels, and is solved by using the graph-cut α-expansion approach. The energy to optimize is computed as a sum of data and interaction energy terms, respectively. The data energy term is computed using the outputs of the probabilistic support vector machines classification. The second energy term, which expresses the interaction between spatially adjacent pixels, is computed by using dissimilarity measures between spectral vectors, such as vector norms, spectral angle mapper and spectral information divergence. Experimental results on hyperspectral images captured by the RO-SIS and the AVIRIS sensors reveal that the proposed method yields higher classification accuracies when compared to the recent state-of-the-art approaches.

Today, water quality and water pollution can be detected using remote sensing and its terrestrial components. Remote sensing does not only provide a quick solution to detect water quality and pollution, but it could also be low cost. Within the scope of the study, the water quality of the İmranlı area of the Kızılırmak River, one of the most important water resources of the Sivas region and the İmranlı dam on the river, was investigated by spectral classification methods. Water samples were taken from various points on the river and dam and their chemical oxygen demands were determined in the laboratory. In addition, the reflectance values of the water samples taken by the local spectral measurements were examined in order to use as end members for spectral classification. CHRIS Proba is used as satellite image. Match filtering (MF), spectral angle mapping (SAM) and spectral information divergence (SID) methods have been used as the spectral classification methods and it has been examined which method gives better results in determining water quality. According to the results, it is understood that SAM method provides better classification accuracy than other methods.

This study aims at analysing the spectral characters of high turbid coastal waters to retrieve the Suspended Sediment Concentration (SSC) in Gulf of Cambay, India. For this purpose, the Ocean Colour Monitor (OCM) data was processed using quasi-synchronous field, and laboratory spectral reflectance measurements. Spectral similarity between the image-spectra and library-spectra estimated by Spectral Angle Mapper (SAM) algorithm was used to retrieve SSC. The result obtained by SAM algorithm was compared with Total Suspended Matter (TSM) and Suspended Sediment Matter Retrieval (SSMR) algorithms. It was found that for case-II waters, the longer wavelength (743 nm and 835 nm) is very sensitive to SSC changes. Hence, the SAM-based approach involving all wavelength regions is found to be very efficient and match with field measured SSC.