Pietro Guccione

Hyperbolic RDM for Nonlinear Valumetric Distortions

Quantization-based watermarking techniques are sensitive to valumetric scaling, a wide class of distortions applied to images and videos, such as contrast change or gamma correction. Several methods have been proposed to counteract valumetric attacks, but the common approach to the problem is to take only the linear ones into account. This paper presents an extension to the rational dither modulation (RDM) data-hiding scheme which provides robustness against nonlinear distortions modelled by a power-law attack. The algorithm makes use of proper mapping of the pixel values from the Cartesian to hyperbolic coordinates. This mapping is able to render the problem similar to the classical RDM scheme, since fixed multiplicative scaling is cancelled out while the exponentiation of a nonlinear distortion is transformed into a gain scaling. The validity of the approach has been confirmed by applying the watermarking scheme to Gaussian host and real images; experimental results confirm its intrinsic invariance against the power-law attack. Finally, it will be shown that under the white Gaussian noise addition, the proposed scheme achieves a good bit-error rate (BER). The measured BER is affected by the properties of the embedding domain, as supported by the theoretical analysis given in this paper.

Flexible Dynamic Block Adaptive Quantization for Sentinel-1 SAR Missions

The letter introduces a novel quantizer suited for medium to high-resolution synthetic aperture radar (SAR) systems, like the forthcoming SENTINEL-1 SAR. The Flexible Dynamic Block Adaptive Quantization (FDBAQ) extends the concept of the Block Adaptive Quantization (BAQ), used in spaceborne SAR since the Magellan mission, by adaptively tuning the quantizer rate according to the local signal-to-noise-ratio (SNR). A design is presented aiming to optimize the average bit-rate, while constraining the minimum SNR. FDBAQ optimized performance is then evaluated using backscatter maps derived from ENVIronment SATellite (ENVISAT) data.

Interferometry with ENVISAT wide swath ScanSAR data

The possibility to get efficient topographic mapping and monitoring of large-scale motions with ScanSAR interferometry has been demonstrated with the Shuttle Radar Topography Mission and RADARSAT mission. The Environmental Satellite Advanced Synthetic Aperture Radar (ASAR) sensor has been designed to provide enhanced capabilities for interferometric applications. Different types of interferometric products can be obtained by combining the various ASAR modes as stripmap synthetic aperture radar [image mode (IM)] and ScanSAR [wide swath (WS) mode]. This letter deals with the possibility to use WS data to get either mixed-mode (IM/WS) or ScanSAR mode (WS/WS) differential interferograms. The impact of digital elevation model localization errors on IM/WS interferograms and of scan pattern synchronization on WS/WS interferograms is investigated. Experimental results are encouraging and show that ASAR ScanSAR data can be routinely used for interferometric applications in both cases

Doppler centroid estimation for ScanSAR data

We introduce a novel accurate technique to estimate the Doppler centroid (DC) in ScanSAR missions. The technique starts from the ambiguous DC measures in the subswaths and uses a method alternative to standard unwrapping to undo the jumps in estimates induced by modulo pulse repetition frequency (PRF) measures. The proposed alternative is less error prone than the usual unwrapping techniques. Doppler Ambiguity is then solved by implementing a maximum-likelihood estimate that exploits the different PRFs used in different subswaths. An azimuth pointing of the antenna that does not change with subswaths, or that changes in a known way, is assumed. However, if the PRF diversity is strong enough, unknown small changes in azimuth pointing are tolerated and accurately estimated. This estimator is much simpler and more efficient, than those in the literature. Results achieved with both RADARSAT 1 and ENVISAT ScanSAR data are reported.

Dealing with temporal and spatial correlations to classify outliers in geophysical data streams

Anomaly detection and change analysis are challenging tasks in stream data mining. We illustrate a novel method that addresses both these tasks in geophysical applications. The method is designed for numeric data routinely sampled through a sensor network. It extends the traditional time series forecasting theory by accounting for the spatial information of geophysical data. In particular, a forecasting model is computed incrementally by accounting for the temporal correlation of data which exhibit a spatial correlation in the recent past. For each sensor the observed value is compared to its spatial-aware forecast, in order to identify the outliers. Finally, the spatial correlation of outliers is analyzed, in order to classify changes and reduce the number of false anomalies. The performance of the presented method is evaluated in both artificial and real data streams.

Using trend clusters for spatiotemporal interpolation of missing data in a sensor network

Ubiquitous sensor stations continuously measure several geophysical fields over large zones and long (potentially unbounded) periods of time. However, observations can never cover every location nor every time. In addition, due to its huge volume, the data producedcannot be entirelyrecordedfor futureanalysis. In this scenario, interpolation, i.e., the estimation of unknown data in each location or time of interest, can be used to supple- ment station records. Although in GIScience there has been a tendency to treat space and time separately, integrating space and time could yield better results than treating them separately when interpolating geophysical fields. According to this idea, a spatiotemporal interpolation process, which accounts for both space and time, is described here. It oper- ates in two phases. First, the exploration phase addresses the problem of interaction. This phase is performed on-line using data recordedfrom a network throughout a time window. The trend cluster discovery process determines prominent data trends and geographically- aware station interactions in the window. The result of this process is given before a new data window is recorded. Second, the estimation phase uses the inverse distance weighting approach both to approximate observed data and to estimate missing data. The proposed technique has been evaluated using two large real climate sensor networks. The experi- ments empirically demonstrate that, in spite of a notable reduction in the volume of data, the technique guarantees accurate estimation of missing data.

Iterative Hyperspectral Image Classification Using Spectral–Spatial Relational Features

This paper describes the principles and implementation of an algorithm for the classification of hyperspectral remote sensing images. The proposed approach is novel and can be included within the category of the spectral-spatial classification algorithms. The elements of novelty of the algorithm are as follows: (1) the implementation of two classifiers that work iteratively, each one exploiting the decision of the other to improve the training phase, and (2) the use of relational features based on the current labeling and on the spatial structure of the image. The two classifiers are fed with the spectral features and with the spatial features, respectively. The spatial features are built using the relative abundance of each class in a neighborhood of the pixel (homogeneity index), where the neighborhood is properly defined. An important contribution to the success of the method is the adoption of a multiclass classifier, the multinomial logistic regression, and a proper use of the posterior probabilities to infer the class labeling and build the relational data. The results of the two classifiers are eventually combined by means of an ensemble decision. The algorithm has been successfully tested on three standard hyperspectral images taken from the Airborne Visible-Infrared Imaging Spectrometer and ROSIS airborne sensors and compared with classification algorithms recently proposed in the literature.

A novel spectral-spatial co-training algorithm for the transductive classification of hyperspectral imagery data

Abstract The automatic classificationr of hyperspectral data is made complex by several factors, such as the high cost of true sample labeling coupled with the high number of spectral bands, as well as the spatial correlation of the spectral signature. In this paper, a transductive collective classifier is proposed for dealing with all these factors in hyperspectral image classification. The transductive inference paradigm allows us to reduce the inference error for the given set of unlabeled data, as sparsely labeled pixels are learned by accounting for both labeled and unlabeled information. The collective inference paradigm allows us to manage the spatial correlation between spectral responses of neighboring pixels, as interacting pixels are labeled simultaneously. In particular, the innovative contribution of this study includes: (1) the design of an application-specific co-training schema to use both spectral information and spatial information, iteratively extracted at the object (set of pixels) level via collective inference; (2) the formulation of a spatial-aware example selection schema that accounts for the spatial correlation of predicted labels to augment training sets during iterative learning and (3) the investigation of a diversity class criterion that allows us to speed-up co-training classification. Experimental results validate the accuracy and efficiency of the proposed spectral-spatial, collective, co-training strategy.

Tailored multivariate analysis for modulated enhanced diffraction

Modulated enhanced diffraction (MED) is a technique allowing the dynamic structural characterization of crystalline materials subjected to an external stimulus, which is particularly suited for in situ and operando structural investigations at synchrotron sources. Contributions from the (active) part of the crystal system that varies synchronously with the stimulus can be extracted by an offline analysis, which can only be applied in the case of periodic stimuli and linear system responses. In this paper a new decomposition approach based on multivariate analysis is proposed. The standard principal component analysis (PCA) is adapted to treat MED data: specific figures of merit based on their scores and loadings are found, and the directions of the principal components obtained by PCA are modified to maximize such figures of merit. As a result, a general method to decompose MED data, called optimum constrained components rotation (OCCR), is developed, which produces very precise results on simulated data, even in the case of nonperiodic stimuli and/or nonlinear responses. The multivariate analysis approach is able to supply in one shot both the diffraction pattern related to the active atoms (through the OCCR loadings) and the time dependence of the system response (through the OCCR scores). When applied to real data, OCCR was able to supply only the latter information, as the former was hindered by changes in abundances of different crystal phases, which occurred besides structural variations in the specific case considered. To develop a decomposition procedure able to cope with this combined effect represents the next challenge in MED analysis.

Azimuth Antenna Maximum Likelihood Estimation by Persistent Point Scatterers in SAR Images

This paper addresses the problem of estimating the azimuth antenna pattern by using a set of persistent point scatterers (PPS) retrieved from a stack of interferometric synthetic aperture radar images. This is achieved by means of a maximum likelihood estimation. PPS emerge as a restricted subset of the well known persistent scatterers, for which many applications have been described in the literature. PPS have a more stringent property since they explicitly require an impulsive trend feature; a good degree of isolation from the neighboring targets is further necessary to estimate the antenna pattern by means of digital spotlight focusing. A statistical model for PPS is provided and experimentally validated; the sufficient number of PPSs necessary to get a given accuracy for the azimuth antenna estimation is also suggested. Results using both simulated and real X-band Cosmo Skymed data are eventually illustrated.