In recent years, neural networks have been a key advance in the field of image compression. However, although their performance is highly competitive with established conventional methods, their high computational cost limits their use in settings where available power is a significant limitation, such as satellites and other remote sensing platforms. This problem has been addressed by recent contributions that have proposed reduced-complexity neural image codecs based on compressive autoencoders with hyperprior. These codecs allow for joint compression and denoising, variable-rate or fixed-quality compression. This paper proposes a comparison of the theoretical complexity of several neural codecs, as well as their encoding runtime and power consumption on low-power devices. This study includes comparisons between fixed-rate, precise rate allocation, and fixed-quality compression modes. The viability of their deployment on remote sensing missions is discussed based on these results.

Satellite optical images, upon their on-ground receipt, offer a distorted view of the observed scene. Their restoration, including denoising, deblurring, and sometimes super-resolution, is required before their exploitation. Moreover, quantifying the uncertainties related to this restoration helps to reduce the risks of misinterpreting the image content. Deep learning methods are now state-of-the-art for satellite image restoration. Among them, direct inversion methods train a specific network for each sensor, and generally provide a point estimation of the restored image without the associated uncertainties. Alternatively, deep regularization (DR) methods learn a deep prior on target images before plugging it, as the regularization term, into a model-based optimization scheme. This allows for restoring images from several sensors with a single network and possibly for estimating associated uncertainties. In this paper, we introduce VBLE-xz, a DR method that solves the inverse problem in the latent space of a variational compressive autoencoder (CAE). We adapt the regularization strength by modulating the bitrate of the trained CAE with a training-free approach. Then, VBLE-xz estimates relevant uncertainties jointly in the latent and in the image spaces by sampling an explicit posterior estimated within variational inference. This enables fast posterior sampling, unlike state-of-the-art DR methods that use Markov chains or diffusion-based approaches. We conduct a comprehensive set of experiments on very high-resolution simulated and real Pléiades images, asserting the performance, robustness and scalability of the proposed method. They demonstrate that VBLE-xz represents a compelling alternative to direct inversion methods when uncertainty quantification is required.

Regularization of inverse problems is of paramount importance in computational imaging. The ability of neural networks to learn efficient image representations has been recently exploited to design powerful data-driven regularizers. While state-of-the-art plug-and-play (PnP) methods rely on an implicit regularization provided by neural denoisers, alternative Bayesian approaches consider Maximum A Posteriori (MAP) estimation in the latent space of a generative model, thus with an explicit regularization. However, state-of-the-art deep generative models require a huge amount of training data compared to denoisers. Besides, their complexity hampers the optimization involved in latent MAP derivation. In this work, we first propose to use compressive autoencoders instead. These networks, which can be seen as variational autoencoders with a flexible latent prior, are smaller and easier to train than state-of-the-art generative models. As a second contribution, we introduce the Variational Bayes Latent Estimation (VBLE) algorithm, which performs latent estimation within the framework of variational inference. Thanks to a simple yet efficient parameterization of the variational posterior, VBLE allows for fast and easy (approximate) posterior sampling. Experimental results on image datasets BSD and FFHQ demonstrate that VBLE reaches similar performance as state-of-the-art PnP methods, while being able to quantify uncertainties significantly faster than other existing posterior sampling techniques. The code associated to this paper is available in https://github.com/MaudBqrd/VBLE.

One of the main limitations to the adoption of deep learning for image compression is the need to train multiple models to compress at multiple rates. In the case of onboard remote sensing data compression, another limitation is the computational cost of the neural networks. Addressing both limitations, this paper presents a new reduced-complexity architecture for multi-rate compression of remote sensing images. The proposed architecture enables compressing at a precise user-selected rate while keeping a competitive performance in lossy compression on different sets of remote sensing data. The proposed approach is amenable for onboard deployment.

Earth observation through satellite images is crucial to help economic activities as well as to monitor the impact of human activities on ecosystems. Current satellite systems are subjected to strong computational complexity constraints. Thus, image compression is performed onboard with specifically tailored algorithms while image denoising is performed on the ground. In this letter, we intend to address satellite image compression and denoising with neural networks. The first proposed approach uses a single neural architecture for joint onboard compression and denoising. The second proposed approach sequentially uses a first neural architecture for onboard compression and a second one for on ground denoising. For both approaches, the onboard architectures are lightened as much as possible, following the procedure proposed by Alves de Oliveiraet al.(2021). The two approaches are shown to outperform the current satellite imaging system and their respective pros and cons are discussed.

Archetypal scenarios for change detection generally consider two images acquired through sensors of the same modality. The resolution dissimilarity is often bypassed though a simple preprocessing, applied independently on each image to bring them to the same resolution. However, in some important situations, e.g. a natural disaster, the only images available may be those acquired through sensors of different modalities and resolutions. Therefore, it is mandatory to develop general and robust methods able to deal with this unfavorable situation. This paper proposes a coupled dictionary learning strategy to detect changes between two images with different modalities and possibly different spatial and/or spectral resolutions. The pair of observed images is modelled as a sparse linear combination of atoms belonging to a pair of coupled overcomplete dictionaries learnt from the two observed images. Codes are expected to be globally similar for areas not affected by the changes while, in some spatially sparse locations, they are expected to be different. Change detection is then envisioned as an inverse problem, namely estimation of a dual code such that the difference between the estimated codes associated with each image exhibits spatial sparsity. A comparison with state-of-the-art change detection methods evidences the proposed method superiority.

This paper deals with a fusion-based change detection (CD) framework for multi-band images with different spatial and spectral resolutions. The first step of the considered CD framework consists in fusing the two observed images. The resulting fused image is subsequently spatially or spectrally degraded to produce two pseudo-observed images, with the same resolutions as the two observed images. Finally, CD can be performed through a pixel-wise comparison of the pseudo-observed and observed images since they share the same resolutions. Obviously, fusion is a key step in this framework. Thus, this paper proposes to quantitatively and qualitatively compare state-of-the-art fusion methods, gathered into four main families, namely component substitution, multi-resolution analysis, unmixing and Bayesian, with respect to the performance of the whole CD framework evaluated on simulated and real images.

Change detection (CD) is one of the most challenging issues when analyzing remotely sensed images. Comparing several multidate images acquired through the same kind of sensor is the most common scenario. Conversely, designing robust, flexible, and scalable algorithms for CD becomes even more challenging when the images have been acquired by two different kinds of sensors. This situation arises in the case of emergency under critical constraints. This paper presents, to the best of our knowledge, the first strategy to deal with optical images characterized by dissimilar spatial and spectral resolutions. Typical considered scenarios include CD between panchromatic, multispectral, and hyperspectral images. The proposed strategy consists of a three-step procedure: 1) inferring a high spatial and spectral resolution image by fusion of the two observed images characterized one by a low spatial resolution and the other by a low spectral resolution; 2) predicting two images with, respectively, the same spatial and spectral resolutions as the observed images by the degradation of the fused one; and 3) implementing a decision rule to each pair of observed and predicted images characterized by the same spatial and spectral resolutions to identify changes. To quantitatively assess the performance of the method, an experimental protocol is specifically designed, relying on synthetic yet physically plausible change rules applied to real images. The accuracy of the proposed framework is finally illustrated on real images.

This paper proposes a robust fusion-based strategy to detect changes between two multi-band optical images with different spatial and spectral resolutions, e.g., a multispectral high spatial resolution image and a hyperspectral low spatial resolution image. The dissimilarity between sensor resolutions makes the change detection problem challenging, which has been generally bypassed in the literature: most often, the two images are crudely and independently resampled in order to get the same spatial and spectral resolutions and finally, classical change detection methods are applied. However, the resampling operation tends to lose information. In this paper, we propose a method that more effectively uses the available information: the two observed images are respectively modeled as spatial and spectral degradations of two latent images characterized by the same high spatial and high spectral resolutions. Representing the same scene, these latent images are expected to be globally similar except for possible changes in sparse spatial locations. Change detection is then envisioned through the solution of an inverse problem, shown to be a specific instance of multi-band image fusion. The proposed method is applied to real images with simulated realistic changes. A comparison with state-of-the-art change detection methods evidences the proposed method superiority.

Increasing the sampling rate of Analog-to-Digital Converters (ADC) is a main challenge in many fields and especially in telecommunications. Time-Interleaved ADCs (TI-ADC) were introduced as a technical solution to reach high sampling rates by time interleaving and multiplexing several low-rate ADCs at the price of a perfect synchronization between them. Indeed, as the signal reconstruction formulas are derived under the assumption of uniform sampling, a desynchronization between the elementary ADCs must be compensated upstream with an online calibration and expensive hardware corrections of the sampling device. Based on the observation that desynchronized TI-ADCs can be effectively modeled using a Periodic Non-uniform Sampling (PNS) scheme, we develop a general method to blindly estimate the time delays involved in PNS. The proposed strategy exploits the signal stationarity properties and thus is simple and quite generalizable to other applications. Moreover, contrarily to state-of-the-art methods, it applies to bandpass signals which is the more judicious application framework of the PNS scheme.

In recent years, remote sensing of the Earth surface using images acquired from aircraft or satellites has gained a lot of attention. The acquisition technology has been evolving fast and, as a consequence, many different kinds of sensors (e.g., optical, radar, multispectral, and hyperspectral) are now available to capture different features of the observed scene. One of the main objectives of remote sensing is to monitor changes on the Earth surface. Change detection has been thoroughly studied in the case of images acquired by the same sensors (mainly optical or radar sensors). However, due to the diversity and complementarity of the images, change detection between images acquired with different kinds of sensors (sometimes referred to as heterogeneous sensors) is clearly an interesting problem. A statistical model and a change detection strategy were recently introduced in [J. Prendes, M. Chabert, F. Pascal, A. Giros, and J.-Y. Tourneret, Proceedings of the IEEE International Conference on Acoustics, Speech and Signal Processing, Florence, Italy, 2014; IEEE Trans. Image Process., 24 (2015), pp. 799--812] to deal with images captured by heterogeneous sensors. The main idea of the suggested strategy was to model the objects contained in an analysis window by mixtures of distributions. The manifold defined by these mixtures was then learned using training data belonging to unchanged areas. The changes were finally detected by thresholding an appropriate distance to the estimated manifold. This paper goes a step further by introducing a Bayesian nonparametric framework allowing us to deal with an unknown number of objects in analysis windows without specifying an upper bound for this number. A Markov random field is also introduced to account for the spatial correlation between neighboring pixels. The proposed change detector is validated using different sets of synthetic and real images (including pairs of optical images and pairs of optical and radar images) showing a significant improvement when compared to existing algorithms.

This paper introduces a Bayesian non parametric (BNP) model associated with a Markov random field (MRF) for detecting changes between remote sensing images acquired by homogeneous or heterogeneous sensors. The proposed model is built for an analysis window which takes advantage of the spatial information via an MRF. The model does not require any a priori knowledge about the number of objects contained in the window thanks to the BNP framework. The change detection strategy can be divided into two steps. First, the segmentation of the two images is performed using a region based approach. Second, the joint statistical properties of the objects in the two images allows an appropriate manifold to be defined. This manifold describes the relationships between the different sensor responses to the observed scene and can be learnt from a training unchanged area. It allows us to build a similarity measure between the images that can be used in many applications such as change detection or image registration. Simulation results conducted on synthetic and real optical and synthetic aperture radar (SAR) images show the efficiency of the proposed method for change detection.

Remote sensing images are commonly used to monitor the earth surface evolution. This surveillance can be conducted by detecting changes between images acquired at different times and possibly by different kinds of sensors. A representative case is when an optical image of a given area is available and a new image is acquired in an emergency situation (resulting from a natural disaster for instance) by a radar satellite. In such a case, images with heterogeneous properties have to be compared for change detection. This paper proposes a new approach for similarity measurement between images acquired by heterogeneous sensors. The approach exploits the considered sensor physical properties and specially the associated measurement noise models and local joint distributions. These properties are inferred through manifold learning. The resulting similarity measure has been successfully applied to detect changes between many kinds of images, including pairs of optical images and pairs of optical-radar images.

This paper introduces a new statistical model for homogeneous images acquired by the same kind of sensor (e.g., two optical images) and heterogeneous images acquired by different sensors (e.g., optical and synthetic aperture radar (SAR) images). The proposed model assumes that each image pixel is distributed according to a mixture of multi-dimensional distributions depending on the noise properties and on the transformation between the actual scene and the image intensities. The parameters of this new model can be estimated by the classical expectation-maximization algorithm. The estimated parameters are finally used to learn the relationships between the different images. This information can be used in many image processing applications, particularly those requiring a similarity measure (e.g., change detection or registration). Simulation results on synthetic and real images show the potential of the proposed model. A brief application to change detection between optical and SAR images is finally investigated.

This paper proposes a method that simultaneously builds the analytic signal from non-uniform samples of a bandpass signal and rejects interferences. The analytic signal is required for many onboard operations in communication satellites. This method operates in the time domain and without preliminary demodulation, using Periodic Non-uniform Sampling of order 2 (PNS2). This non-uniform sampling scheme can be easily implemented with available devices. Exact formulas for the analytic signal construction are derived for an infinite observation window (an infinite number of samples). For practical applications, the formulas should also demonstrate a high convergence rate due to the finite observation window. Formulas with increasing convergence rates are thus derived. The proposed method has been tested through simulations according to the number of available samples, the interference parameters and the filter transfer function regularity.

Analysis and comparison of linear and hybrid/compact dual-polarization (dual-pol) synthetic aperture radar (SAR) imagery have gained a wholly new importance in the last few years, in particular, with the advent of new spaceborne SARs such as the Japanese ALOS PALSAR, the Canadian RADARSAT-2, and the German TerraSAR-X. Compact polarimetry, hybrid dual-pol, and quad-pol modes are newly promoted in the literature for future SAR missions. In this paper, we investigate and compare different hybrid/compact and linear dual-pol modes in terms of the estimation of the degree of polarization (DoP). The DoP has long been recognized as one of the most important parameters characterizing a partially polarized electromagnetic wave. It can be effectively used to characterize the information content of SAR data. We study and compare the information content of the intensity data provided by different hybrid/compact and linear dual-pol SAR modes. For this purpose, we derive the joint distribution of multilook SAR intensity images. We use this distribution to derive the maximum likelihood and moment-based estimators of the DoP in hybrid/compact and linear dual-pol modes. We evaluate and compare the performance of these estimators for different modes on both synthetic and real data, which are acquired by RADARSAT-2 spaceborne and NASA/JPL airborne SAR systems, over various terrain types such as urban, vegetation, and ocean.

This paper studies the reconstruction of a random band-limited process from a nonuniform periodic sampling of order 2. This sampling scheme uses two interleaved sample sequences at the same rate chosen according to the Landau criterion. The associated exact reconstruction formula is composed of two series with cardinal sine functions as general terms. In practical applications, the signal is observed on a finite window. The exact reconstruction formula is then approached by truncation of the series. The series rate of convergence has thus a strong influence on the reconstruction performance in practice. Unfortunately, the cardinal sine functions lead to a slow convergence. In the case of an over-sampled process, we propose interpolation formulas derived from raised cosine filters. We show, through theoretical analysis and numerical simulations, that the regularity of these functions leads to a higher series convergence rate and thus improves the reconstruction performance even for a low over-sampling rate.

This paper deals with non uniform sampling of remote sensing optical images. Non uniform sampling can model defects of the uniform sampling device such as grid distortion or dead pixels. Moreover, non uniform sampling can be intentionally introduced either for its good properties with respect to aliasing or in the case of variable rate devices such as telemetry and remote control. Image reconstruction from non uniform samples requires appropriate interpolation formulas taking into account the sample locations. This paper compares cubic spline interpolation and a modified Lagrangian formula in terms of Peak Signal to Noise Ratio (PSNR) in the case of high resolution remote sensing optical images. Non uniform sampling shows a higher robustness with respect to aliasing than the classical uniform sampling. In case of non uniform sampling, the modified Lagrange reconstruction formula leads to a higher image quality than the cubic spline interpolation.

This paper studies the potential interest of the multivariate Pearson system for the statistical modeling of images acquired by different sensors. These images include optical and synthetic aperture radar (SAR) remote sensing images. The univariate Pearson system has shown good capacities to capture the statistical properties of SAR images. This paper introduces a generalization of this system to multidimensional vectors with a particular attention to the bivariate case. The method of moments is investigated to estimate the unknown parameters of bivariate Pearson distributions. The estimation performance is evaluated using synthetic and real data.

Synthetic Aperture Radar (SAR) is a powerful tool strongly employed in maritime monitoring and surveillance. Among different polarimetric SAR modes, dual-pol SAR data are widely used for monitoring large ocean and coastal areas. The degree of polarization (DoP) is a fundamental quantity characterizing a partially polarized electromagnetic field. The performance of the DoP is studied for ship detection under different polarizations in hybrid/compact and linear dual-pol SAR imagery. Experiments are performed on C-band polarimetric data acquired by RADARSAT-2 over San Francisco Bay and the Strait of Gibraltar.

Information contained in polarimetric images can be characterized by a scalar parameter called the polarization degree. This parameter is usually estimated using four polarimetric images. However, acquisition and registration of four images are complex and costly. Thus, reducing the number of required images can be of great interest for practical applications. In coherent illumination, these images are degraded by speckle noise. This noise can be reduced by transforming the original single-look images into multi-look images. We provide maximum likelihood estimators of the polarization degree in the general case of multi-look intensity images. The estimators are derived based on only two measurements, under coherent illumination and fully developed speckle. We evaluate our method on synthetic data and compare its performance with that of moment-based methods.

Analysis of dual-polarized SAR imagery has gained new importance with the recent launches of ALOS PALSAR, RADARSAT-2, and TerraSAR-X polarimetric SAR systems. Information contained in polarimetric images, collected by these SAR systems, can be characterized by a scalar parameter called the degree of polarization. This parameter has long been recognized as one of the most important parameters characterizing a partially polarized electromagnetic wave. In this paper, we provide maximum likelihood and moment-based estimators of the degree of polarization in dual-polarized SAR imagery. We evaluate and compare the performance of these estimators on RADARSAT-2 polarimetric data, over various terrain types such as urban, vegetation, and ocean.

This paper studies database updating using optical and synthetic aperture radar images. Logistic regression is used to model the conditional probability of presence/absence of buildings given features extracted from the images. The logistic regression parameters are estimated using the maximum likelihood method. Binary hypothesis tests are then constructed from these estimates to detect changes between the optical/radar images and the existing database. The estimation and detection algorithms are evaluated using simulated and real data sets.

The degree of polarization (DoP) has long been recognized as one of the most important parameters characterizing partially polarized electromagnetic waves. This parameter can be effectively used to describe the information content of polarimetric images collected by synthetic aperture radar (SAR) systems. Estimation of DoP is standardly performed using four measurements. In SAR compact polarimetry (CP), however, only two measurements are available. In this paper, we develop maximum likelihood estimators of the DoP, in SAR CP modes, based on only two intensity images. We evaluate and compare the performance of these estimators for different CP modes on RADARSAT-2 polarimetric data, over various terrain types such as urban, vegetation, and ocean.

This paper addresses the problem of defining a similarity measure between an observed Gaussian image and a binary image constructed from a cartographic database. The main idea is to assume that the binary image has been obtained by thresholding an unobserved Gaussian image correlated with the observed image. The proposed statistical model is then used to estimate its unknown parameters using the maximum likelihood method. The paper discusses a possible application to change detection between a cartographic vector data base and an optical image.

This paper presents a satellite image compression scheme based on a post-processing of the wavelet transform of images. The bandelet transform is a directional post-processing of wavelet coefficients. Thanks to a low computational complexity, this transform is a good candidate for future on-board satellite image compression systems. First, we analyze the ability of the bandelets to exploit directional correlations between wavelet coefficients. This study leads to an improved post-processing with a better decorrelation of adjacent wavelet coefficients in the vertical or in the horizontal direction taking into account the wavelet subband orientations. To perform even better decorrelation, bases are also build by principal component analysis (PCA). This results in an improved compression performance without increasing the computational complexity.

This paper studies the performance of the ZC contexts in JPEG2000. A quality criterion based on the mutual information is introduced. According to this criterion, there exist ZC contexts which are not effective in any bit-plane. They can thus be merged with others without significant loss of the compression efficiency. Since the contextual conditional probabilities strongly depend on the bit-planes, an adaptation of the contexts to the bit-planes is also proposed. First, an optimal context quantizer is used on each bit-planes, then a simple greedy entropy minimization procedure allows to combine the ZC contexts in JPEG2000. The results are compared with a global combination of JPEG2000 ZC contexts. Tests have been performed on a real satellite image data base. With only four ZC contexts, the PSNR is less than 0.01 dB lower than the PSNR with the nine ZC contexts.

In this paper, linear periodically time varying (LPTV) filters are used as an alternative to direct sequence modulation in spread spectrum watermarking techniques. Due to their periodicity, LPTV filters combined with a Peano-Hilbert scan in the spatial domain of natural images offer good decoding performance. Specific LPTV filters are proposed to reject the low-pass host noise and to shape the watermark spectrum. Extensions to improved spread spectrum and quantization in a subspace projection are also investigated.

This paper proposes a new method for mechanical fault detection in induction motors. The detection strategy is based on the estimation of a particular stator current parameter. The considered mechanical faults cause periodic load torque oscillations leading to a sinusoidal phase modulation of the stator current. The modulation index is related to the fault severity and can be used as a fault indicator. First, a simplified stator current model is proposed. The problem is then equivalent to the parameter estimation of a sinusoidal phase mono-component signal. Second, the maximum likelihood estimator is implemented using evolution strategies for optimization. The Cramer-Rao lower bounds are calculated and compared to the estimator performance through simulations. The estimation procedure is studied on experimental stator current signals from faulty and healthy motors

Interpolation often acts as a perturbation in watermarking schemes. In an opposite approach, this article proposes a watermarking algorithm based on interpolation in the spatial domain. The perceptual properties of interpolation allow to generate an imperceptible mark and to substitute this mark to the host image. A theoretical study of the detection and decoding performance is provided, as well as the robustness to attacks compared to classical watermarking algorithms. Imperceptibility and security properties of this scheme are also discussed

This paper addresses abrupt change detection in multiplicative noise using the continuous wavelet transform. An optimal wavelet, maximizing a well-chosen time-scale contrast criterion is derived. The analytical optimization gives the optimal wavelet closed expression. The influence of the mother wavelet on signature-based detector performance is then demonstrated. Detection performance is characterized using receiver operating characteristic curves computed from Monte-Carlo simulations. The optimal wavelet obviously improves performance with respect to other wavelets classically used for singularity detection.

We propose sliding-window multiedge detectors and reflectivity estimators for complex SAR images. The novel detectors and estimators allow to take into account additive observation noise and colored signal (speckle) and noise processes; furthermore, they employ an exponential data weighting to improve spatial resolution. In the multiedge case, simulation results demonstrate a substantial performance improvement over existing methods when the speckle is colored and additive noise is present.

The problem addressed in the paper is the detection of abrupt changes embedded in multiplicative colored Gaussian noise. The multiplicative noise is modeled by an AR process. The Neyman Pearson detector is developed when the abrupt change and noise parameters are known. This detector constitutes a reference to which suboptimal detectors can be compared. In practical applications, the abrupt change and noise parameters have to be estimated. The maximum likelihood estimator for these parameters is then derived. This allows to study the generalized likelihood ratio detector.

This paper addresses the problem of additive and multiplicative abrupt jump detection. Two detection algorithms based on the continuous wavelet transform are studied. Additive and multiplicative jumps are compared in the time-scale plane. Considering non-zero mean signals, we show that the same wavelet based algorithms can be used for the detection of additive and multiplicative jumps. However, an analysis of the variance in the time-scale plane allows us to distinguish these two kinds of jump.