Accurate Local Climate Zone (LCZ) classification is essential for urban climate studies, environmental monitoring, and sustainable city planning. Recent advances in Deep Learning (DL) have significantly improved LCZ mapping, but challenges remain in capturing spatial position features and distinguishing spectrally similar land cover types. This letter proposes multi-scale Coordinate Attention-based Multi-Stream Local Climate Zone Network (MSCA-MSLCZNet), a multi-stream DL framework integrating multi-scale feature processing, attention mechanisms, and rule-based refinement to enhance LCZ classification performance. The model is evaluated on the So2Sat LCZ42 dataset, outperforming baseline methods in overall accuracy (OA), OA of built-up classes (OAbu), OA of natural classes (OAn), and Kappa. Further validation on Milan LCZ mapping confirms its generalization capability, demonstrating strong classification performance in built-up areas and improved urban structure delineation. Comparative experiments highlight the model’s ability to better differentiate urban structures and built-up zones from natural landscapes. MSCA-MSLCZNet proves effective for large-scale LCZ mapping, offering improved classification accuracy and adaptability to diverse geographic regions.

Abstract—Coastal water quality monitoring is crucial for environmental sustainability and public health. This work introduces a very cutting-edge methodology, using ΦSat-2 multispectral data and quanvolutional neural networks to explore quantum-enhanced machine learning for water contaminant assessment. By integrating quantum preprocessing into a classical regression model, it is possible to achieve a significant reduction in model parameters while maintaining high predictive accuracy. Additionally, this work introduces an innovative dataset that integrates simulated ΦSat-2 spectral data with Copernicus Marine Service bio-geochemical products, ensuring a strong alignment between satellite observations and reference turbidity measurements. Our results show that quantum models use up to 98% fewer parameters than their classical counterparts, while achieving a 6.9% improvement in the Pearson correlation coefficient between the ΦSat-2 pre-processed bands and the ground-truth turbidity values, compared to the case without quantum pre-processing. Additionally, the Root Mean Square Error (RMSE) improves by 7.3% over the classical baseline. These findings highlight the potential of quantum-assisted remote sensing to enable more efficient and scalable analysis of large-scale water contaminant data, paving the way for advanced big data approaches in water quality monitoring.

Detecting earthquake-induced landslides in remote sensing images is challenging due to the varying sizes of landslides, uneven distribution, and the prevalence of small targets. This study proposes a novel approach, the TLSTMF-YOLO model, which combines a C3-Swin-Transformer and multiscale feature fusion techniques to enhance detection accuracy and efficiency. Key innovations include the use of a convolutional block attention module (CBAM) to improve feature representation, and a bidirectional feature pyramid network (BiFPN) for optimized cross-scale feature fusion. To address data scarcity, a transfer learning strategy is applied, supported by an AdamW optimizer and cosine learning rate strategy for faster convergence. Evaluations on the Jiuzhaigou and Luding landslide datasets demonstrate the model’s effectiveness, achieving precision, recall, and mean average precision (mAP)@0.5 of 95.7%, 89.9%, and 90.5% on the Jiuzhaigou dataset, and 96.0%, 90.9%, and 94.5% on the Luding dataset, respectively. In addition, the model processes frames efficiently, with times of 6.61 and 12.2 ms on the two datasets. These results confirm the model’s capability for accurate and efficient landslide detection, highlighting its potential for real-world applications.

A significant amount of remotely sensed data is generated daily by many Earth observation (EO) spaceborne and airborne sensors over different countries of our planet. Different applications use those data, such as natural hazard monitoring, global climate change, urban planning, and more. Many challenges are brought by the use of these big data in the context of remote sensing (RS) applications. In recent years, the employment of machine learning (ML) and deep learning (DL)-based algorithms has allowed a more efficient use of these data, but the issues in managing, processing, and efficiently exploiting them have even increased as classical computers have reached their limits. This article highlights a significant shift toward leveraging quantum computing (QC) techniques in processing large volumes of RS data. The proposed Quanv4EO framework introduces a quanvolution method for (pre)processing multidimensional EO data. Its effectiveness was first demonstrated on standard image classification datasets (MNIST and FashionMNIST), achieving accuracies of 99.84% and 96.81%, respectively, with a significantly reduced model size of 42 k parameters and 16 frozen qubits. Its capabilities were then checked on EO datasets, such as EuroSAT, with a mean accuracy of 96% using balanced iterative reducing and clustering using hierarchies (BIRCHs) clustering and 93% using automated DL (AutoDL), surpassing or matching state-of-the-art (SOTA) classical nonquantum models. Applying the framework to synthetic aperture radar (SAR) data, the QSPeckleFilter demonstrates notable improvements in speckle noise reduction, achieving a peak signal-to-noise ratio (PSNR) of 21.72 and a structural similarity index measure (SSIM) of 0.81, surpassing all tested classical counterparts. The proposed results underscore the potential of quantum-enhanced approaches in RS data analysis, paving the way for more efficient and effective solutions for wide geographical area EO data exploitation.

The use of Synthetic Aperture Radar (SAR) has greatly advanced our capacity for comprehensive Earth monitoring, providing detailed insights into terrestrial surface use and cover regardless of weather conditions, and at any time of day or night. However, SAR imagery quality is often compromised by speckle, a granular disturbance that poses challenges in producing accurate results without suitable data processing. In this context, the present paper explores the cutting-edge application of Quantum Machine Learning (QML) in speckle filtering, harnessing quantum algorithms to address computational complexities. We introduce here QSpeckleFilter, a novel QML model for SAR speckle filtering. The proposed method compared to a previous work from the same authors showcases its superior performance in terms of Peak Signal-to-Noise Ratio (PSNR) and Structural Similarity Index Measure (SSIM) on a testing dataset, and it opens new avenues for Earth Observation (EO) applications.

This paper explores an innovative fusion of Quantum Computing (QC) and Artificial Intelligence (AI) through the development of a Hybrid Quantum Graph Convolutional Neural Network (HQGCNN), combining a Graph Convolutional Neural Network (GCNN) with a Quantum Multilayer Perceptron (MLP). The study highlights the potentialities of GCNNs in handling global-scale dependencies and proposes the HQGCNN for predicting complex phenomena such as the Oceanic Niño Index (ONI). Preliminary results suggest the model potential to surpass state-of-the-art (SOTA). The code will be made available with the paper publication.

Segmentation of buildings in urban areas is crucial in many applications such as urban planning, disaster response, and population mapping. The size and resolution of satellite images as well as the density of urban areas add challenges to building segmentation. This study proposes using a U-Net architecture as a Deep Learning (DL) approach to segment buildings in urban areas using Very High Resolution (VHR) COSMO-SkyMed images which represents a true novelty with respect to the state of the art (SOTA). A GitHub page has been created to make available code and dataset. The outcomes of the proposed method are very promising and future works will be also discussed in the end. The code is available at github.com/BabakMemar/buildingSegmentation-UNet.

Monitoring water contaminants is of paramount importance, ensuring public health and environmental well-being. Turbidity, a key parameter, poses a significant problem, affecting water quality. Its accurate assessment is crucial for safeguarding ecosystems and human consumption, demanding meticulous attention and action. For this, our study pioneers a novel approach to monitor the Turbidity contaminant, integrating CatBoost Machine Learning (ML) with high-resolution data from Sentinel-2 Level-2A. Traditional methods are labor-intensive while CatBoost offers an efficient solution, excelling in predictive accuracy. Leveraging atmospherically corrected Sentinel-2 data through the Google Earth Engine (GEE), our study contributes to scalable and precise Turbidity monitoring. A specific tabular dataset derived from Hong Kong contaminants monitoring stations enriches our study, providing region-specific insights. Results showcase the viability of this integrated approach, laying the foundation for adopting advanced techniques in global water quality management.

Climate change is intensifying extreme weather events, causing both water scarcity and severe rainfall unpredictability, and posing threats to sustainable development, biodiversity, and access to water and sanitation. This paper aims to provide valuable insights for comprehensive water resource monitoring under diverse meteorological conditions. An extension of the SEN2DWATER dataset is proposed to enhance its capabilities for water basin segmentation. Through the integration of temporally and spatially aligned radar information from Sentinel-1 data with the existing multispectral Sentinel-2 data, a novel multisource and multitemporal dataset is generated. Benchmarking the enhanced dataset involves the application of indices such as the Soil Water Index (SWI) and Normalized Difference Water Index (NDWI), along with an unsupervised Machine Learning (ML) classifier (k-means clustering). Promising results are obtained and potential future developments and applications arising from this research are also explored.

In remote sensing, target detection in hyperspectral systems is a crucial duty since it enables the localization and discrimination of target features. For this purpose, reflectance spectra are frequently utilized, and the spectral signatures with corresponding component abundances in the observed scene are displayed. Nevertheless, many hyperspectral sensors have restricted spatial resolution, namely only a part of the pixel is occupied by the targets, and the spectra of multiple sub-pixel targets, along with the background spectrum, gets combined within a single pixel. Therefore, we propose in this paper a generalized replacement model that considers different sub-pixel target spectra and execute the detection process as a binary hypothesis test. This method shows to work well in handling this problem.

With artificial intelligence breakthroughs permeating the Earth Science domain, there is an immediate need to advance the data, tools, and resulting technologies to broader societal challenges. Different efforts are emerging with fragmented best practices for making Earth Observation (EO) data Artificial Intelligence (AI)-ready, availing computer vision and image analysis tools for broader reuse across the remote sensing community. This paper will revisit current best practices and outline a guideline for advancing EO data and derivative AI products for broader community use. We mainly discuss the Analysis Ready Data (ARD) essentials and aim to forge their evolution with Findable, Accessible, Interoperable, Reusable (FAIR) principles to support cross-modal/cross-sensor/cross-provider opportunities that appear to be central to solving complex EO challenges.

Climate change has caused disruption in certain weather patterns, leading to extreme weather events like flooding and drought in different parts of the world. In this paper, we propose machine learning methods for analyzing changes in water resources over a time period of six years, by focusing on lakes and rivers in Italy and Spain. Additionally, we release open-access code to enable the expansion of the study to any region of the world. We create a novel multi-spectral and multitemporal dataset, SEN2DWATER, which is freely accessible on GitHub. We introduce suitable indices to monitor changes in water resources, and benchmark the new dataset on three different deep learning frameworks: Convolutional Long Short Term Memory (ConvLSTM), Bidirectional ConvLSTM, and Time Distributed Convolutional Neural Networks (TD-CNNs). Future work exploring the many potential applications of this research is also discussed.

Climate change and increases in drought conditions affect the lives of many and are closely tied to global agricultural output and livestock production. This research presents a novel approach utilizing machine learning frameworks for drought prediction around water basins. Our method focuses on the next-frame prediction of the Normalized Difference Drought Index (NDDI) by leveraging the recently developed SEN2DWATER database. We propose and compare two prediction methods for estimating NDDI values over a specific land area. Our work makes possible proactive measures that can ensure adequate water access for drought-affected communities and sustainable agriculture practices by implementing a proof-of-concept of short and long-term drought prediction of changes in water resources.

This work presents an unsupervised Deep Learning (DL)-based method for the matching of SAR images in case of high variability of the data. The method is based on an autoencoder trained in an unsupervised manner, with the SAR image in input and output given at the same time. From this architecture, the encoder is detached and used as feature extractor to obtain a set of feature maps at different depths to be used for the comparison. The correlation is applied to these feature maps to get the point of maximum correspondence among the image and the template to be found. The DL model has been trained with a heterogeneous dataset, with images coming from different SAR sources, in order to make it generalised and to make it work in a wide range of scenarios. The method has been tested both with similar and completely different images, showing good performances even in the most critical case.

Quantum Machine Learning (QML) is an emerging technology that only recently has begun to take root in the research fields of Earth Observation (EO) and Remote Sensing (RS), and whose state of the art is roughly divided into one group oriented to fully quantum solutions, and in another oriented to hybrid solutions. Very few works applied QML to EO tasks, and none of them explored a methodology able to give guidelines on the hyperparameter tuning of the quantum part for Land Cover Classification (LCC). As a first step in the direction of quantum advantage for RS data classification, this letter opens new research lines, allowing us to demonstrate that there are more convenient solutions to simply increasing the number of qubits in the quantum part. To pave the first steps for researchers interested in the above, the structure of a new hybrid quantum neural network for EO data and LCC is proposed with a strategy to choose the number of qubits to find the most efficient combination in terms of both system complexity and results accuracy. We sampled and tried a number of configurations, and using the suggested method we came up with the most efficient solution (in terms of the selected metrics). Better performance is achieved with less model complexity when tested and compared with state-of-the-art (SOTA) and standard techniques for identifying volcanic eruptions chosen as a case study. Additionally, the method makes the model more resilient to dataset imbalance, a significant problem when training classical models. Lastly, the code is freely available so that interested researchers can reproduce and extend the results.

In this paper, an automatic processing chain, fully de-veloped in C++, for the generation of simplified Synthetic Aperture Radar (SAR) images of large urban scenarios is presented. The proposed method makes use of open GIS data and an open-source simulation technique based on ray tracing. The main novelties refer to: 1) the integration of a method for generating Three-Dimensional (3D) models of almost any arbitrary place on Earth, including in the model land cover information of the scene, 2) the possibility of setting the SAR parameters for the creation of images acquired under different geometries (e.g. satel-lite and airborne with arbitrary view (or look) angles), and 3) the ability to have at the end of the processing chain a geo-referenced image for comparisons with other types of data. The authors strongly believe that this pa-per can be of wide interest to researchers in the field of Remote Sensing (RS).

In this paper, the authors aim to design a decision support system (DSS) based on machine learning (ML) to assist institutions in implementing targeted countermeasures to combat and prevent emergencies such as the COVID -19 pandemic. The DSS relies on an ensemble of several ML models that combine heterogeneous data to predict risk levels at the micro and macro levels. Some preliminary analyses have already been conducted showing the corre-lation between nitrogen dioxide (N0O), mobility-related parameters, and COVID -19 data. However, given the complexity of the virus spread mechanism, which is re-lated to many different factors, these preliminary stud-ies confirmed the need to perform more in-depth analyses on the one hand and to use ML algorithms on the other hand to capture the hidden relationships between the huge amounts of data that need to be processed.

Cloud removal is a relevant topic in Remote Sensing, fostering medium- and high-resolution optical image usability for Earth monitoring and study. Recent applications of deep generative models and sequence-to-sequence-based models have proved their capability to advance the field significantly. Nevertheless, there are still some gaps: the amount of cloud coverage, the landscape temporal changes, and the density and thickness of clouds need further investigation. We fill some of these gaps in this work by introducing an innovative deep model. The proposed model is multi-modal, relying on both spatial and temporal sources of information to restore the whole optical scene of interest. We use the outcomes of both temporal-sequence blending and direct translation from Synthetic Aperture Radar (SAR) to optical images to obtain a pixel-wise restoration of the whole scene. The reconstructed images preserve scene details without resorting to a considerable portion of a clean image. Our approach’s advantage is demonstrated across various atmospheric conditions tested on different datasets. Quantitative and qualitative results prove that the proposed method obtains cloud-free images coping with landscape changes.

The goal of this project is to investigate global nightlights and light pollution as indicators for the economic growth and urbanization of a nation. We test this proposition using the nightlight data obtained by the Defense Meteorological Satellite Program (DMSP) of the National Oceanic and Atmospheric Administration (NOAA). Through examining the changes over time of spatial nightlight patterns for different developing countries on Google Earth Engine (GEE), we investigate how they correlate to countries' development indicators such as Gross Domestic Product (GDP) and other parameters which represent new reference indicators with respect to previous research in the literature. A preliminary analysis brings to interesting results and discussion.

This article aims to explore the potential of current approaches for quantum image classification in the context of remote sensing. After a brief outline of quantum computers and an analysis of the current bottlenecks, it shows for the first time experiments with quantum neural networks on a reference Earth observation (EO) dataset: EuroSAT. Moreover, it establishes the proof of concept of quantum computing for EO: the models trained and run on a quantum simulator are on par with classical ones. We make the open-source code available for further developments11QNN4EO repository: https://github.com/ESA-PhiLab/QNN4EO..

This paper presents an unsupervised workflow to assess the extension of the burned areas due to wildfire events over green vegetated areas, in particular woods and forests, based on open access and high spatial resolution multispectral images from the Landsat-8 and Sentinel-2 satellites, together with open access high temporal resolution products (MODIS and VIIRS). The idea is to use jointly these kinds of data to combine the advantages linked to the use of high spatial resolution data with those ones derived from the exploitation of high temporal resolution data. Experimental results, carried on different case studies, show the effectiveness of the proposed method on forest fires, by reaching a performance in terms of the F-measure parameter, that ranges from 0.69 to 0.95, depending on the specific case study.

In this paper the authors present and validate a procedure for bridge monitoring, based on freely available satellite data and the straightforward Differential SAR Interferometry (DInSAR) technique. A displacement dataset of the Morandi bridge in Genoa (Italy) has been created, before its collapse. The outputs obtained were then compared to those found in literature and achieved through the Persistent Scatterer Interferometry (PSI), a more complex and reliable technique. Results demonstrate that the adopted procedure has great potentiality in the application field and could effectively be extended to different types of civil infrastructures.

In this paper, an in-depth study is presented regarding an innovative methodology on which the authors have invested several months of research, as evidenced by their recent references on similar topics. A semi-automatic classification of the buildings in urban area is analyzed, when Light Detection And Ranging (LiDAR) data are combined to Very High Resolution (VHR) WorldView-2 (WV-2) images and an Object-Based Image Analysis (OBIA) technique is used. The aim of the data fusion from different sensor types is to realize an effective and limited-cost product for protection and management of environmental and natural resources, to meet the needs of many institutions, such as municipalities, provinces and regions, and for applications in the context of risk analysis, territorial planning and local development. This procedure may be extended to large areas of the territory, resulting into a reduction of the processing time for the building detection, with respect to a human visual or manual photointerpretation. Interesting results and further implications will be presented and discussed.

In this paper, the authors aim to combine the latest state of the art models in image recognition with the best publicly available satellite images to create a system for landslide risk mitigation. We focus first on landslide detection and further propose a similar system to be used for prediction. Such models are valuable as they could easily be scaled up to provide data for hazard evaluation, as satellite imagery becomes increasingly available. The goal is to use satellite images and correlated data to enrich the public repository of data and guide disaster relief efforts for locating precise areas where landslides have occurred. Different image augmentation methods are used to increase diversity in the chosen dataset and create more robust classification. The resulting outputs are then fed into variants of 3-D convolutional neural networks. A review of the current literature indicates there is no research using CNNs (Convolutional Neural Networks) and freely available satellite imagery for classifying landslide risk. The model has shown to be ultimately able to achieve a significantly better than baseline accuracy.

This paper deals with ground displacement measurements with Differential Synthetic Aperture Radar Interferometry (DInSAR) technique. These ground modifications often occur as a consequence of an earthquake. The island of Ischia (Southern Italy) has been chosen as case study since it was hit by a severe earthquake on the 21stof August 2017. The National Institute of Geophysics and Volcanology (INGV) and the Institute for Electromagnetic Sensing of the Environment (IREA) of the National Research Council of Italy (CNR) provided the displacement maps considering interferometric pairs close to the main shock event. In this work, a further interferometric pair, which also includes some aftershocks, has been used to calculate ground modifications. The results confirmed a ground subsidence up to 4 centimeters in the epicenter area, in agreement with the reference data. This study rely on open access data as well as on open software, both provided by the European Space Agency (ESA) under the Copernicus program. In particular, data from Sentinel-1 radar satellite mission and SNAP software have been used. The availability of such data and software is relevant for public institutions that can produce valuable information at almost no charge.

In this paper the Two-Scale Model has been used to derive the theoretical Normalized Radar Cross Section (NRCS) for sea clutter in the L-band, taking into account also the circular polarization of GPS signals. Using this theoretical model and the theoretical formula for NRCS, authors aim to investigate the possibility to extend target detection through GPS signals in backscattering configuration by varying the incidence angle and the wind speed, whereas results obtained previously were related only to a single value of the incidence angle and to two values of the wind speed. Target Scattered Power and Sea Clutter Power are derived and compared finding that: 1) smaller targets can be detected as the incidence angle increases; 2) the Sea Clutter Power is lower than the one estimated when the “worst case” of a VV-polarization signal is chosen to derive the NRCS value from experimental data. Some final considerations are made as future work.

Authors in this work aim to present new analysis methods for Earth Observation, developed by processing Sentinel-1 and Landsat-8 satellite data and combining them in an original way. Comparing SAR and Optical/Multispectral data is a procedure already in use because they are two acquisition systems that provide very different and therefore complementary and useful information. Even if the combination of such different data is not a simple process, the overall information greatly improves when both satellite data are jointly used, as application of our procedure to some case studies demonstrates.

A new approach for simulation of delay-Doppler maps for ocean global navigation satellite system reflectometry is presented. The simulator is based on a stochastic forward scattering model for generation of realizations of the sea surface reflected signal and allows a flexible definition of system and environmental parameters. The signal scattered from a delay Doppler cell is modelled as the coherent random sum of a large number of components from specular points; the sum converges to a compound-Gaussian random variable when the mean number of points is sufficiently large. In the proposed simulation process, the discretization of the DDM arises from a natural sampling of the delay-Doppler domain rather than on sampling the ocean surface. Instantaneous DDMs are simulated and consistency with the theoretical model is verified.

In this work, authors outline how measurements from different types of sensors can be put together to characterize in an extensive way a sensitive waste dump site. Terrestrial surveying systems and aerial images have been combined with satellite remote sensing techniques to get a complete overview of the surface stability and its thermal characteristics in conjunction with geomorphological conditions and water distribution. The objective has been to implement a broad monitoring system to control the area under observation. Interesting observations have been deducted from the comparison among different data since surface temperature retrievals and variations can be correlated to soil moisture and water exchange by showing that the proposed monitoring system performs correctly in taking under control deformation, terrain displacements, temperature and water surface modifications.

The Nadir reflectances, measured by the SCIAMACHY hyperspectral sensor, are used to retrieve vertical column concentrations of nitrogen dioxide (NO2), resulting from anthropogenic pollution. The estimation process is realized via a blind source separation method: the unmixing of the NO2spectral waveform from the overall atmospheric absorption contribution within the logarithmic reflectance spectra is realized.

Atmospheric concentrations of trace-gases are retrieved from hyperspectral data using a blind source separation method. The algorithm relies on the assumption that the absorption cross sections of the gas components are weakly dependent on the overall atmospheric background. The unmixing of contributions from the logarithm of the spectral reflectance provides estimates of both individual trace-gas absorption cross sections and their concentrations. In the experimental analysis, nadir reflectances received by SCanning Imaging Absorption spectroMeter for Atmospheric CHartographY are considered in two scenarios: the sulfur dioxide emissions from a volcanic eruption and the nitrogen dioxide production from anthropogenic pollution. In both cases, it is demonstrated that the algorithm performs very similarly to the Differential Optical Absorption Spectroscopy algorithm but with very little ancillary information.

The average density of specular points plays an important role in the investigation of the sea surface features with remote sensing systems. Evidently, the number of specular reflectors has a direct physical link with the surface roughness that is, in turn, modified by the wind velocity, by the presence of internal waves, oil films and geometric modifications of the sea surface. The main attempt of this paper is to provide an accurate model for the statistics of the number of wind-induced specular points on the sea surface whose spatial correlation properties are provided by the Elfouhaily sea-spectrum model and the temporal correlation is obtaind by introducing a birth-death-immigration process for the number of specular points. Although the model can be also considered for other applications, it is here applied to GNSS-reflectometry where the small scale roughness can be neglected and time decorrelation is given by the motion of the sea waves and by the navigation of the satellite platforms.

Main goal of this work has been to evaluate the accuracy of digital surface models (DEMs) extracted from in-track EROS-B high resolution satellite stereo imagery. The ability to capture high-resolution stereoscopic images can produce DEMs that are useful in different applications, especially environmental monitoring and mapping. This paper illustrates the results achieved during an experimental campaign for monitoring the Montaguto landslide area in Campania Region, Italy, through the use of stereo EROS-B imagery. The extraction of DEMs has been made using two commercial software: Leica Photogrammetry Suite (LPS) and PCI Geomatica, equipped with the orbital model of EROS-B satellite. The extracted DEMs were compared with those obtained by a LiDAR (Light Detection And Ranging) system and through classical aerial photogrammetry. Both software packages have highlighted an excellent quality in DEM extraction from EROS-B panchromatic images. Good results and an overall satisfactory performance are shown when a comparison with LiDAR data and the Regional Technical Chart (CTR) is done in terms of altitude differences and topographic profiles.

A high-contrast inpainting scheme, based on the Complex Ginzburg-Landau equation, is successfully applied to restoration of SAR interferograms. The algorithm demonstrates quite effective in recovering the phase values in low coherence regions. The validation setup has been carried out in the presence of additive phase noise with a suitable probability density function. Results show that the application of the proposed algorithm to small regions produces very good results both in terms of Signal-to-Noise Ratio (SNR) and in terms of Mean Square Error (MSE).

A high-contrast inpainting scheme based on the Complex Ginzburg-Landau equation recently applied successfully to image restoration is applied to SAR interferograms to improve their quality and therefore final quality of Digital Elevation Models (DEMs). The new technique attempts to recover the phase values in low coherence regions through diffusion and inpainting. After phase unwrapping low coherence regions are masked and discarded and a Complex Ginzburg-Landau (CGL) inpainting scheme is applied to regions where phase values are missing. We demonstrate that the residues reduce and the proposed algorithm leads to a higher Signal-to-Noise Ratio (SNR) if compared with MCF algorithm. The restoration technique has been applied to ERS-1 and ERS-2 data sets acquired on July 1995. Results appear to be very promising: the proposed algorithm provides good performances especially in presence of strong noise level and low coherence areas with relatively small dimensions.

Multispectral sensors using array of detectors are affected by striping, an artifact that appears as a series of horizontal bright or dark periodic lines in the remotely sensed images. Nonlinearities and memory effect of detectors are the main causes of the striping problem that is not effectively corrected in the onboard or postprocessing calibration phases. In order to clear striping from images, we consider a new procedure based on detector response equalization and apply it to moderate resolution imaging spectroradiometer data from Terra and Aqua satellites. After identification of theout-of-familydetectors, a least squares equalization stage is considered for calibration by using the intrinsic data redundancy caused by the bow-tie effect, where multiple observations of the same field of view are available fr.om different detectors. The main advantage of this method, with respect to others such as the histogram equalization, is due to the independence of the measurements on the scene statistics, which, otherwise, will cause an overestimation or underestimation of the detectors' responses. The new procedure performance is validated using data received at the Mediterranean Agency for Remote Sensing and environmental control ground station facility in benevento-Italy and data downloaded from NASA LAADS Web site. The main results are presented, by showing the effectiveness of the method and the stability of the correction coefficients, at least on one-orbit periods.

An image registration algorithm based on mutual information maximization and non-linear diffusion is presented. It relies on a non-parametric estimation of the degree of dependency between the reference image and the template to be registered, which is intrinsically more robust against possible deformations due to imaging geometry and propagation disturbances. The approach based on non linear diffusion, nonetheless, has the advantage of producing a non-parametric discrete warping model which does not rely on a particular set of basis functions, and is therefore as much general as possible. The experimental results on simulated images have quantitatively shown the accuracy of the proposed method.

Hybrid sensor networks (HSNs), made up of many static sensors interacting with a limited number of mobile robotic sensors, can support a variety of applications in which assisted navigation, implemented for mobile robotic sensors, performs requested operations in hostile environments, where the human presence is impracticable. On the other hand, the limited on-board battery energy of the sensor nodes imposes hard constraints to the wireless communication and computing capabilities of such nodes, thus limiting the capacity of a HSN to run an application for a period of interest without human intervention. These considerations have motivated this paper, which describes a method to estimate the energy consumption under changeable working conditions in a HSN running an assisted navigation algorithm based on the "credit field" concept.

This paper deals with the problem of CFAR detection of thermal anomalies in multispectral satellite data. The goal is to extend the algorithm proposed in [1], and successfully applied to MODIS data from band 21, to the case of multiband investigation. A multiple-channel model has been designed, where data from MODIS bands 21 and 31 are projected into a new coordinates system by adopting the Principal Component Analysis (PCA). A preliminary statistical analysis has been performed on both the principal components of data to verify that the Weibull distribution can be adopted for background. Subsequently, a Kendall test has been used to check the level of dependency of the projected data and it has shown that channels independence can be assumed with high significance level. After PCA, a CFAR detection is applied to projected data and thanks to data independence the single detections are combined with an AND rule. The outcome of the AND operation gives the thermal anomalies detected in both channels with an assigned overall probability of false alarm (PFA). The Multiband CFAR algorithm has been applied to a 256 × 256 MODIS image from bands 21 and 31 and results have been compared with those from NASA-DAAC MOD14.

In hybrid sensor networks (HSNs) one or more mobile sensor nodes can interact with other mobile sensor nodes or static ones in order to collaboratively perform specific tasks. Such networks can support applications in which assisted navigation is implemented for mobile sensor nodes that carry out requested operations through a monitored environment. To this end, several assisted navigation algorithms have been proposed within the HSNs scope with the goal of overcoming the typical problems of classic sensor networks (SNs), such as limited transmission power and lean autonomy in terms of computing and memory capacity. Nevertheless, the distributed and highly dynamic nature of such algorithms still makes their implementation in a real scenario a difficult task. In this paper we present an improved version of the assisted navigation algorithm based on the "credit field" concept. The algorithm has been designed for HSNs and implemented in Agilla, a middleware running on top of TinyOS and supporting a programming paradigm based on mobile agents. The implementation has been tested on a HSN that employs Mica2 Motes as static sensor nodes and the Lego Mindstorms robot integrated with a Stargate node developed by Crossbow as a mobile sensor node.

Abstract — This paper introduces the concept of Constant False Alarm Rate (CFAR) in fire detection for multispectral satellite data. A new algorithm is proposed, based on a technique successfully applied for detection of extended objects in High Resolution SAR images. It compares the pixel under analysis with an adaptive threshold, suitably estimated from the pixels surrounding the one under test, in order to ensure the CFAR property. The proposed approach requires that the background distribution is of Location Scale (LS) type or amenable to such a distribution by a suitable transformation. MODIS data from the 4 µm channel are considered. A preliminary statistical analysis is performed to verify if the Weibull distribution, compliant with LS representation, can be adopted for background. MODIS cloud and water masking are applied to identify those pixels to be discarded before implementing the statistical analysis. Results of fire detection are presented for different values of the system parameters (censoring depth and false alarm rate) and compared with the algorithm implemented in the NASA-DAAC MOD14. I.