Single-polarization Synthetic Aperture Radar Images Research Articles

A Group-Wise Feature Enhancement-and-Fusion Network with Dual-Polarization Feature Enrichment for SAR Ship Detection

Ship detection in synthetic aperture radar (SAR) images is a significant and challenging task. However, most existing deep learning-based SAR ship detection approaches are confined to single-polarization SAR images and fail to leverage dual-polarization characteristics, which increases the difficulty of further improving the detection performance. One problem that requires a solution is how to make full use of the dual-polarization characteristics and how to excavate polarization features using the ship detection network. To tackle the problem, we propose a group-wise feature enhancement-and-fusion network with dual-polarization feature enrichment (GWFEF-Net) for better dual-polarization SAR ship detection. GWFEF-Net offers four contributions: (1) dual-polarization feature enrichment (DFE) for enriching the feature library and suppressing clutter interferences to facilitate feature extraction; (2) group-wise feature enhancement (GFE) for enhancing each polarization semantic feature to highlight each polarization feature region; (3) group-wise feature fusion (GFF) for fusing multi-scale polarization features to realize polarization features’ group-wise information interaction; (4) hybrid pooling channel attention (HPCA) for channel modeling to balance each polarization feature’s contribution. We conduct sufficient ablation studies to verify the effectiveness of each contribution. Extensive experiments on the Sentinel-1 dual-polarization SAR ship dataset demonstrate the superior performance of GWFEF-Net, with 94.18% in average precision (AP), compared with the other ten competitive methods. Specifically, GWFEF-Net can yield a 2.51% AP improvement compared with the second-best method.

Flood Detection in Dual-Polarization SAR Images Based on Multi-Scale Deeplab Model

The proliferation of massive polarimetric Synthetic Aperture Radar (SAR) data helps promote the development of SAR image interpretation. Due to the advantages of powerful feature extraction capability and strong adaptability for different tasks, deep learning has been adopted in the work of SAR image interpretation and has achieved good results. However, most deep learning methods only employ single-polarization SAR images and ignore the water features embedded in multi-polarization SAR images. To fully exploit the dual-polarization SAR data and multi-scale features of SAR images, an effective flood detection method for SAR images is proposed in this paper. In the proposed flood detection method, a powerful Multi-Scale Deeplab (MS-Deeplab) model is constructed based on the dual-channel MobileNetV2 backbone and the classic DeeplabV3+ architecture to improve the ability of water feature extraction in SAR images. Firstly, the dual-channel feature extraction backbone based on the lightweight MobileNetV2 separately trains the dual-polarization SAR images, and the obtained training parameters are merged with the linear weighting to fuse dual-polarization water features. Given the multi-scale space information in SAR images, then, a multi-scale feature fusion module is introduced to effectively utilize multi-layer features and contextual information, which enhances the representation of water features. Finally, a joint loss function is constructed based on cross-entropy and a dice coefficient to deal with the imbalanced categorical distribution in the training dataset. The experimental results on the time series of Sentinel-1A SAR images show that the proposed method for flood detection has a strong ability to locate water boundaries and tiny water bodies in complex scenes. In terms of quantitative assessment, MS-Deeplab can achieve a better performance compared with other mainstream semantic segmentation models, including PSPNet, Unet and the original DeeplabV3+ model, with a 3.27% intersection over union (IoU) and 1.69% pixel accuracy (PA) improvement than the original DeeplabV3+ model.

Fast Task-Specific Region Merging for SAR Image Segmentation

In existing superpixel-wise segmentation algorithms, superpixel generation most often is an isolated preprocessing step. The segmentation performance is determined to a certain extent by the accuracy of superpixels. However, it is still a challenge to develop a stable superpixel generation method. In this article, we attempt to incorporate the superpixel generation and merging steps into an end-to-end trainable deep network. First, we employ a recently proposed differentiable superpixel generation method to over-segment the single-polarization synthetic aperture radar (SAR) image. It outputs the statistical likelihood that each pixel belongs to different superpixels. In superpixel merging part, as one of our main contributions, we propose a superpixel-wise statistical dissimilarity measure method for converting the soft superpixels set into a self-connected weighted graph. More importantly, inspired by the concept of the number of walks in graph theory, we define the <inline-formula> <tex-math notation="LaTeX">$k$ </tex-math></inline-formula>-order connectivity of each vertex. This definition can intelligently indicate the potential soft cluster centers and class assignments in graph. This merging method is differentiable, computationally simple, and free of empirical parameters. The superpixel generation and merging phases can be implemented under a unified deep network. The benefit is that our method can iteratively adjust the shapes of the superpixels according to the boundaries and segmentation results during training, until the satisfactory segmentation results are captured. Experimental results on real SAR images demonstrate that the segmentation precision of our proposed method is superior to other state-of-the-art methods in terms of precision and computational efficiency.

Ship Segmentation via Encoder–Decoder Network With Global Attention in High-Resolution SAR Images

Ship detection in the synthetic aperture radar (SAR) image is of great significance in the fields of military and coastal defense. Most ship detection methods are designed based on the object detection framework, which can only provide the vertices’ coordinates of the bounding box covering the ship targets but cannot provide more detailed contour information. Target segmentation can further explore the shape and edge information of the objects, which can be used as a blazing novel means for automatic object detection. In this letter, a 3-D atrous encoder–decoder neural network with global attention modules (GAM-EDNet) is proposed to achieve ship segmentation in SAR images. The encoder–decoder structure with atrous convolution is developed as the network body to fully exploit the structural information of the ship targets with various sizes. To increase the structural information of the single-polarization SAR images, a 3-D image cube is designed as the input of the GAM-EDNet. A global attention module is proposed to further improve the segmentation performance by integrating the high-level semantic features with the low-level location features. Besides, an SAR ship segmentation dataset (SAR-HR4) is built to evaluate the segmentation performance, and the experimental results show that the proposed GAM-EDNet achieves better performance than other state-of-the-art methods.

Fast SAR Image Segmentation With Deep Task-Specific Superpixel Sampling and Soft Graph Convolution

Since the number of superpixels is lower than that of pixels, superpixels can substantially speed up subsequent processing steps and have been widely used in synthetic aperture radar (SAR) image segmentation. However, in most of the existing superpixel-wise segmentation algorithms, superpixel prediction is an isolated preprocessing step and is independent of the segmentation task. The performance of the segmentation results is determined by the accuracy of superpixels. Once superpixels are generated, their shape cannot be changed in the following segmentation stage, even if the same superpixels contain pixels of different landcovers. To address this, we propose an end-to-end trainable superpixel-wise segmentation method for single-polarization SAR images. First, we design a differentiable boundary-ware clustering method for estimating task-specific superpixels. Instead of the hard association between pixels and superpixels in the existing superpixel algorithms, this method introduces the soft association map to make the clustering differentiable. Hence, it can be implemented using a simple deep fully convolutional network. In the segmentation part, we propose a novel soft graph convolution network (Soft-GCN), which takes the association map as input and performs superpixel-wise segmentation. The advantage of our method is that superpixel generation and graph convolution parts can be trained under a unified framework, until two parts obtain the optimum parameters. In the training process, it can adaptively adjust the shape of the superpixels according to the segmentation results, ensuring the superpixels correctly adhere the boundaries. Experimental results with simulated and real SAR images demonstrate that our method outperforms other state-of-the-art segmentation algorithms, while also being faster.

A dual-scale siamese densely connected network with MRF for SAR image classification

ABSTRACT Recently, Convolutional Neural Network (CNN) has achieved some success in synthetic aperture radar (SAR) image classification. This outstanding performance mainly depends on a large number of training samples, but achieved satisfactory classification results with limited training samples remains a challenge. To address this problem, we propose a dual-scale Siamese densely connected network with Markov Random Fields (DS-SDCNet-MRF) for single-polarization SAR image classification. First, the Siamese densely connected network (SDCNet) is proposed to fully extract discriminative features under limited samples. Then, the proposed DS-SDCNet is constructed with two SDCNets of different scales to produce complementary classification results. Among them, the large-scale SDCNet has a better classification result in the homogeneous region, while the small-scale SDCNet tends to provide good detail preservation. Finally, an improved MRF model which combined the category probability information is proposed to further improve the classification performance. Experimental results on simulated and real single-polarization SAR data demonstrate that the proposed method achieves more encouraging classification performance than the current state-of-the-art classification methods with limited samples.

Performance Comparison of Oil Spill and Ship Classification from X-Band Dual- and Single-Polarized SAR Image Using Support Vector Machine, Random Forest, and Deep Neural Network

It is well known that the polarization characteristics in X-band synthetic aperture radar (SAR) image analysis can provide us with additional information for marine target classification and detection. Normally, dual-and single-polarized SAR images are acquired by SAR satellites, and then we must determine how accurate the marine mapping performance from dual-polarized (pol) images is versus the marine mapping performance from the single-pol images in a given machine learning model. The purpose of this study is to compare the performance of single- and dual-pol SAR image classification achieved by the support vector machine (SVM), random forest (RF), and deep neural network (DNN) models. The test image is a TerraSAR-X dual-pol image acquired from the 2007 Kerch Strait oil spill event. For this, 824,026 pixels and 1,648,051 pixels were extracted from the image for the training and test, respectively, and sea, ship, oil, and land objects were classified from the image by using the three machine learning methods. The mean f1-scores of the SVM, RF, and DNN models resulting from the single-pol image were approximately 0.822, 0.882, and 0.889, respectively, and those from the dual-pol image were about 0.852, 0.908, and 0.898, respectively. The performance improvement achieved by dual-pol was about 3.6%, 2.9%, and 1% in SVM, RF, and DNN, respectively. The DNN model had the best performance (0.889) in the single-pol test while the RF model was best (0.908) in the dual-pol test. The performance improvement was approximately 2.1% and not noticeable. If the condition that dual-pol images have two-times lower spatial resolution versus single-pol images in the azimuth direction is considered, a small improvement may not be valuable. Therefore, the results show that the performance improvement by X-band dual-pol image may be not remarkable when classifying the sea, ships, oil spills, and sea and land surfaces.

HDEC-TFA: An Unsupervised Learning Approach for Discovering Physical Scattering Properties of Single-Polarized SAR Image

Understanding the physical properties and scattering mechanisms contributes to synthetic aperture radar (SAR) image interpretation. For single-polarized SAR data, however, it is difficult to extract the physical scattering mechanisms due to lack of polarimetric information. Time-frequency analysis (TFA) on complex-valued SAR image provides extra information in frequency perspective beyond the “image” domain. Based on TFA theory, we propose to generate the subband scattering pattern for every object in complex-valued SAR image as the physical property representation, which reveals backscattering variations along slant-range and azimuth directions. In order to discover the inherent patterns and generate a scattering classification map from single-polarized SAR image, an unsupervised hierarchical deep embedding clustering (HDEC) algorithm based on TFA (HDEC-TFA) is proposed to learn the embedded features and cluster centers simultaneously and hierarchically. The polarimetric analysis result for quad-pol SAR images is applied as reference data of physical scattering mechanisms. In order to compare the scattering classification map obtained from single-polarized SAR data with the physical scattering mechanism result from full-polarized SAR, and to explore the relationship and similarity between them in a quantitative way, an information theory based evaluation method is proposed. We take Gaofen-3 quad-polarized SAR data for experiments, and the results and discussions demonstrate that the proposed method is able to learn valuable scattering properties from single-polarization complex-valued SAR data, and to extract some specific targets as well as polarimetric analysis. At last, we give a promising prospect to future applications.

On the Nonlinear Mapping of an Ocean Wave Spectrum into a New Polarimetric SAR Image Spectrum

AbstractA new nonlinear transformation relation is derived to describe the mapping of a two-dimensional ocean wave spectrum into a new polarimetric synthetic aperture radar (SAR) image spectrum. It is a further expansion and improvement of Hasselmann’s work. First, the nonlinear mapping relation proposed is derived on the basis of a new polarimetric SAR image instead of the conventional single-polarization SAR image. Second, the nonlinear mapping relation no longer includes the complex hydrodynamic modulation transfer function (MTF). Third, the traditional tilt MTF, which is not accurate enough for the retrieval of sea wave spectrum, is replaced by an empirical tilt MTF derived on the basis of the C-band geophysical model function [i.e., C-band synthetic aperture radar (CSAR) normalized radar cross section (NRCS) model]. A sea wave spectrum retrieval algorithm is then proposed that is based on the new nonlinear mapping and the empirical tilt MTF. The retrieved spectra from C-band polarizedRADARSAT-2SAR images are compared with the results obtained by the ECMWF Ocean Wave Model (ECWAM) and buoy measurements.

SAR Image Classification Using Greedy Hierarchical Learning With Unsupervised Stacked CAEs

Synthetic aperture radar (SAR) can provide stable data source for earth observation due to its advantages of all day and night, all-weather, and strong penetration. SAR image classification as a fundamental procedure has been proved its great value in plenty of remote sensing applications. Conventional classification algorithms mainly rely on hand-designed features, which are susceptible to widespread coherent speckle noise and geometric distortion in high-resolution SAR images. Inspired by the recent impressive success in data mining and deep learning, a greedy hierarchical convolutional neural network (GHCNN) is developed. It aims at obtaining optimized feature representation, relieving the effect of speckle noise, and promoting the local pattern recognition of geometric distortion in single-polarized SAR image classification. First, a series of convolutional autoencoders (CAEs) is trained in the greedy layer-wise unsupervised strategy. This step provides an unbiased regularizer and <i>a priori</i> distribution derived from large volumes of unlabeled SAR patches. Then, to optimize multiple parameter subspaces globally, several CAEs are coupled together to form a deeper hierarchical structure in a stacked and unsupervised fashion. Afterward, a convolutional network with identical topology inherits the pretrained weights. After supervised finetuning, it realizes class prediction. Synchronously, t-distributed stochastic neighbor embedding (t-SNE) algorithm is applied to monitor the efficiency of feature representation during the training period. Experimental results demonstrate that the proposed method has competitive advantages over involved contrast methods.

Deep SAR-Net: Learning objects from signals

This paper introduces a novel Synthetic Aperture Radar (SAR) specific deep learning framework for complex-valued SAR images. The conventional deep convolutional neural networks based methods usually take the amplitude information of single-polarization SAR images as the input to learn hierarchical spatial features automatically, which may have difficulties in discriminating objects with similar texture but discriminative scattering patterns. Our novel deep learning framework, Deep SAR-Net, takes complex-valued SAR images into consideration to learn both spatial texture information and backscattering patterns of objects on the ground. On the one hand, we transfer the detected SAR images pre-trained layers to extract spatial features from intensity images. On the other hand, we dig into the Fourier domain to learn physical properties of the objects by joint time-frequency analysis on complex-valued SAR images. We evaluate the effectiveness of Deep SAR-Net on three complex-valued SAR datasets from Sentinel-1 and TerraSAR-X satellite and demonstrate how it works better than conventional deep CNNs, especially on man-made objects classes. The proposed datasets and the trained Deep SAR-Net model with all codes are provided.

Dual and Single Polarized SAR Image Classification Using Compact Convolutional Neural Networks

Accurate land use/land cover classification of synthetic aperture radar (SAR) images plays an important role in environmental, economic, and nature related research areas and applications. When fully polarimetric SAR data is not available, single- or dual-polarization SAR data can also be used whilst posing certain difficulties. For instance, traditional Machine Learning (ML) methods generally focus on finding more discriminative features to overcome the lack of information due to single- or dual-polarimetry. Beside conventional ML approaches, studies proposing deep convolutional neural networks (CNNs) come with limitations and drawbacks such as requirements of massive amounts of data for training and special hardware for implementing complex deep networks. In this study, we propose a systematic approach based on sliding-window classification with compact and adaptive CNNs that can overcome such drawbacks whilst achieving state-of-the-art performance levels for land use/land cover classification. The proposed approach voids the need for feature extraction and selection processes entirely, and perform classification directly over SAR intensity data. Furthermore, unlike deep CNNs, the proposed approach requires neither a dedicated hardware nor a large amount of data with ground-truth labels. The proposed systematic approach is designed to achieve maximum classification accuracy on single and dual-polarized intensity data with minimum human interaction. Moreover, due to its compact configuration, the proposed approach can process such small patches which is not possible with deep learning solutions. This ability significantly improves the details in segmentation masks. An extensive set of experiments over two benchmark SAR datasets confirms the superior classification performance and efficient computational complexity of the proposed approach compared to the competing methods.

Integration of Spectral Histogram and Level Set for Coastline Detection in SAR Images

This study proposed a novel algorithm for coastline detection in single-polarization synthetic aperture radar (SAR) images based on the local spectral histogram (LSH) and the level set method (LSM). The proposed algorithm includes two main steps. In the processing step, a rough land/sea segmentation is done by utilizing a texture-based segmentation using LSH. In the postprocessing step, the region-based LSM is used to refine the previous segmentation and extract the coastline accurately. The level set (LS) function is initialized using the results of LSH segmentation. This prevents the development of spurious contours, becoming trapped in the local minimum, eliminates the manual support, and also speeds up the LS evolution. A hierarchical LS regularization is proposed using two Gaussian kernels, which is compatible with noisy images and able to detect very narrow regions. The proposed algorithm is able to detect a coastline at the full resolution of the input SAR image and is also robust to noise. A criterion to quantify the accuracy of the results was also proposed. The experimental results for a number of real high-resolution single-polarization SAR images demonstrate that the proposed method is robust to noise and efficient for the coastline detection in different coastal and sea environments.

A Lightweight Feature Optimizing Network for Ship Detection in SAR Image

Deep learning-based methods have achieved great success in target detection tasks of computer vision, but when it comes to Synthetic Aperture Radar (SAR) image ship detection, some new challenges appear because of the wide swath of images, diverse appearances of ships and lack of detail information, which make the detection inefficient and less effective. Aiming to these issues, in this paper, a lightweight feature optimizing network (LFO-Net) based on popular single shot detector (SSD) model is proposed for single polarization SAR image ship detection. Firstly, a simpler structure called lightweight single shot detector (LSSD) is designed, which can be trained from scratch and can reduce the training and testing time without accuracy cost. Secondly, a new bi-directional feature fusion module including one semantic aggregation block and one feature reuse block is proposed to improve the performance of multi-scale targets detection by enhancing the features of both low feature layers and high feature layers. Then the features are further optimized by leveraging attention mechanism, which is beneficial to catch the silent information more efficiently. A set of experiments are implemented to verify the effectiveness of the proposed method using the public SAR ship detection dataset (SSDD). The results show that the proposed method has significant advantages in both speed and accuracy, and outperforms other state-of-art methods. Additionally, a test on GF-3 satellite SAR data with multiple modes verifies the generalization performance of this model.

Significant wave height estimation using azimuth cutoff of C-band RADARSAT-2 single-polarization SAR images

This paper proposes two simple models, look-up table (LUT) model and empirical model, to directly retrieve significant wave height (Hs) using synthetic aperture radar (SAR) azimuth cutoff (λc). Both models aim at C-band VV, HH, VH, and HV single-polarization SAR images. The LUT model relates Hs to λc, while the empirical model relates Hs to both λc and SAR range-to-velocity (β). The LUT model coefficients are derived by simulation under different sea states and observation conditions, which depend on incidence angle (θ), wave direction (dw), and β but are independent of polarization. The empirical model coefficients are obtained by fitting the collocated data, which only depend on polarization. To fit empirical model coefficients and validate the two models, C-band RADARSAT-2 fine quad-polarization (VV+HH+VH+HV) single-look complex (SLC) SAR images and collocated buoy data are collected. Retrieved Hs, using Yang model and the two models proposed in this paper from four kinds of polarization SAR data, are compared with buoy Hs. Results show that both LUT and empirical models have the capacity of retrieving Hs from C-band RADARSAT-2 co-polarization SAR data, while Yang model is not suitable for these kinds of SAR data. Moreover, the empirical model is also valid for cross-polarization SAR data showing clear ocean wave stripes.

Urban Land Use Information Extraction Using the Ultrahigh-Resolution Chinese Airborne SAR Imagery

The rapid development of synthetic aperture radar (SAR) sensors results in the acquisition of substantial ultrahigh-resolution SAR images. In this paper, we, for the first time, present three scenes of single-polarization SAR images with decimeter resolution obtained by a millimeter-wave (MMW) Chinese airborne SAR system. An innovative framework based on the complex generalized Gaussian distribution (CGGD) model is proposed to extract land use information from them, and three CGGD parameters, including the shape parameter, the non-Gaussianity parameter, and the noncircularity parameter, are selected to identify different kinds of ground objects. It is shown that these parameters can reveal plentiful land surface information and will be extremely helpful for single-polarization SAR image interpretations. Moreover, a decision tree classifier is built to categorize these images into homogenous natural surfaces, vegetation textures, circular man-made targets, and noncircular man-made targets. Several interesting experiments are implemented, and their results well demonstrate the effectiveness of the new framework.

Land Cover Classification of Quad-Polarimetric RADARSAT-2 SAR Image Based on Modified Subspace Method

Compared with conventional single-polarization synthetic aperture radar (SAR), quad-polarimetric SAR observations provide more information and show potential for land cover classification. In this study, we propose a modified subspace method (MOSM) for quad-polarimetric SAR image classification, which incorporates the advantages of the learning subspace method (LSM), the averaged learning subspace method (ALSM) and the multiple similarity method (MSM). To evaluate the recognition accuracy of the proposed method, we carry out experiments with quad-polarimetric RADARSAT-2 images for the Yellow River Delta area. We compare the results obtained using the MOSM algorithm with those obtained by the traditional supervised Wishart method. According to our analyses, the results obtained by the proposed method show a 10% higher accuracy rate than the results based on the classical supervised Wishart method, and the overall accuracy rate for the proposed method exceeds 87%, which indicates a good performance. Therefore, the modified method is an alternative classification method for quad-polarimetric SAR image.Résumé. Par rapport au radar à synthèse d’ouverture (RSO) monopolarisation traditionnel, les observations RSO entièrement polarimétriques << quad-polarimetric >> fournissent plus d’informations et montrent un potentiel pour la classification de la couverture terrestre. Dans cet article, nous proposons une méthode de sous-espace modifié (MOSM) pour la classification des images RSO entièrement polarimétriques, qui intègre les avantages de la méthode LSM << learning subspace method >>, de la méthode ALSM << averaged learningsubspace method >> et de la méthode MSM << multiple similarity method >>. Pour évaluer la précision de la reconnaissance de la méthode proposée, nous effectuons des expériences avec des images RADARSAT-2 entièrement polarimétriques pour le delta du fleuve Jaune. Nous comparons les résultats obtenus en utilisant l’algorithme MOSM avec ceux obtenus par la méthode supervisée classique de Wishart. Selon nos analyses, les résultats obtenus par la méthode proposée montrent une précision 10% plus élevée que les résultats basés sur la méthode supervisée classique de Wishart et la précision globale de la méthode proposée dépasse 87%, ce qui indique une bonne performance. Par conséquent, la méthode modifiée est une méthode de classification alternative pour les images RSO entièrement polarimétriques.

Unsupervised Change Detection in Multitemporal SAR Images Over Large Urban Areas

Unsupervised change detection in multitemporal single-polarization synthetic aperture radar (SAR) images often involves thresholding of the image change indicator. If one class, which is usually the unchanged class, comprises a disproportionately large part of the scene, the image change indicator may have a unimodal histogram. Image thresholding of such a change indicator is a challenging task. In this paper, we present an automatic and effective approach to the thresholding of the log-ratio change indicator whose histogram may have one mode or more than one mode. A bimodality test is performed to determine whether the histogram of the log-ratio image is unimodal or not. If it has more than one mode, the generalized Kittler and Illingworth thresholding (GKIT) algorithm based on the generalized Gaussian model (GG-GKIT) is used to detect the optimal threshold values. If it is unimodal, the log-ratio image is divided into small regions and a multiscale region selection process is carried out to select regions which are a balanced mixture of unchanged and changed classes. The selected regions are combined to generate a new histogram. The optimal threshold value obtained from the new histogram is then used to separate unchanged pixels from changed pixels in the log-ratio image. Experimental results obtained on multitemporal SAR images of Toronto and Beijing demonstrate the effectiveness of the proposed approach.

Time-Frequency Analysis in High-Resolution SAR Imagery

In this paper, a time-frequency analysis (TFA) is proposed to derive the backscattering properties of each pixel in single-polarization synthetic aperture radar (SAR) images. At high resolution (HR), some backscattering variations which are linked to the scene geometry and the surface property occur during the radar acquisition. TFA permits to retrieve these variations from the synthesized images. The proposed TFA algorithm is based on a sliding bandpass filtering in the Fourier domain, from which a spectrogram featuring the range and azimuth backscattering variations is derived. The spectrograms summarize the physical properties of each pixel. From the spectrogram analysis, four target classes representing the four main kinds of backscattering behaviors observed in SAR images are defined: frequency invariant, range variant, azimuth variant, and 2-D variant. These classes can further be linked to the physical properties of the objects. An original and simple set of five features estimated from spectrograms is proposed to classify point targets into these four classes. A performance assessment of this classification is carried out, using ONERA/RAMSES X-band airborne images acquired over the city of Toulouse, France. A robustness analysis is also conducted, in order to assess the impact of incidence angle and resolution on the classification performance. Finally, results are also given for spaceborne images (TerraSAR-X spotlight images). The physical interpretation developed in airborne case appears to be also valid for metric spaceborne data. After studying the TFA on HR spaceborne images, the tradeoff between HR coupled with TFA and medium resolution coupled with polarimetric analysis is investigated. Actually, TFA represents another way of characterizing the physical mechanisms involved in image formation.

Characterization of Land Cover Types in TerraSAR-X Images by Combined Analysis of Speckle Statistics and Intensity Information

The appearance of objects and surfaces in synthetic aperture radar (SAR) images significantly differs from the human perception of the environment. In addition, the quality of SAR data is degraded by speckle noise, superposing the true radiometric and textural information of the radar image. Hence, the interpretation of SAR images is considered to be more challenging compared to the analysis of optical data. However, in this paper, we demonstrate how information on the local development of speckle can be used for the differentiation of basic land cover (LC) types in a single-polarized SAR image. For that purpose, we specify the speckle characteristics of the following LC types: 1) water; 2) open land (farmland, grassland, bare soil); 3) woodland; and 4) urban area by means of an unsupervised analysis of scatter plots and standardized histograms of the local coefficient of variation. Next, we use this information for the implementation of a straightforward preclassification of single-polarized TerraSAR-X stripmap images by combining information on the local speckle behavior and local backscatter intensity. The output is either provided as a discrete classification or as a color composite image whose bands can be interpreted in terms of a fuzzy classification. The results of this paper show that unsupervised speckle analysis in high-resolution SAR images supplies valuable information for a differentiation of the water, open land, woodland, and urban area LC types. While the color composite image supports the visual interpretation of SAR data, the outcome of the fully automated discrete LC classification procedure represents a valuable preclassification image, showing overall accuracies of 77%-86%.