Landslide is a common geological disaster, and rapid landslide extraction using high-resolution remote sensing imagery (RSI) is of great significance for emergency rescue and damage assessment. In RSI, landslides often have irregular shapes, large-scale variations, and are easily affected by environmental factors. Existing deep learning methods have limited ability in extracting multiscale features, integrating these features effectively, and adapting to complex environments, resulting in models that are not optimized for robustness. To overcome these challenges, this study proposes a spatial context-guided calibration network (SCGC-Net) for multisource remote sensing data. SCGC-Net introduces a novel combination of hybrid multiscale feature extraction, context-aware modulation of landslide characteristics, and a progressive feature calibration fusion strategy, enabling efficient feature extraction, accurate feature integration, and enhanced cross-domain generalization when working with multisource remote sensing data. SCGC-Net was tested on several datasets representing diverse geographical regions and imaging platforms, including the CAS Landslide Dataset (CLD), HR-GLDD, Bijie, and global very-high-resolution landslide mapping (GVLM). Experimental results indicate that SCGC-Net outperforms existing methods across all evaluation metrics and exhibits superior generalization performance in domain adaptation experiments.

High-resolution soil moisture (SM) products are of broad scientific interest and practical application value. The flexibility of estimating soil moisture from Sentinel-1 data has been widely recognized. However, the backscatter coefficient of synthetic aperture radar (SAR) data is heavily influenced by noise during radiation transmission as well as vegetation cover. This directly leads to the inability to accurately achieve high-resolution SM estimation using a single SAR data. In this paper, we proposed a deep learning model that fuses multisource data that estimates the SM in the depth of ~5 cm. The inputs to the model integrated remote sensing data and soil information data. The model was trained and validated on data from in-situ sensors of the international soil moisture network (ISMN). The proposed method achieved a coefficient of determination (R2) of 0.732 and a root mean square error (RMSE) of 0.069. In addition, the validation experiments in Anhui Province, China also demonstrated the effectiveness and robustness of the proposed method with a R2of 0.85.

The changes in urban built-up areas can reflect the process of urban expansion and have significant implications for evaluating sustainable urban development. The use of remote sensing to detect building areas is an important technology for land monitoring and urban management. Compared with optical images, SAR is less affected by interference and is an important data source for extracting building information in cloudy and rainy areas. Currently, high-precision automated extraction of built-up areas from SAR images remains a challenge. To address this issue, this paper proposes a semantic segmentation framework for building area extraction based on HRNet. Using multiple temporal SAR images acquired from Hainan Province as an example, the paper analyzes changes in the built-up areas of the province by combining the results of built-up area extraction over several years.

Earthquakes are catastrophic in terms of damage to buildings. Synthetic aperture radar (SAR) has emerged as an effective tool to respond to seismic hazards. However, pre- and post-event high-resolution data are not always available for the affected areas, and the complex geometric properties of buildings pose a challenge to building damage detection. Therefore, this letter proposes the Dual-Domain Transformer (DDFormer) semantic segmentation model for damaged buildings detection using a single post-earthquake high-resolution SAR image. The difference between intact and collapsed building features is enhanced by adaptive frequency and spatial modules. Taking the 2023 Turkey earthquake as an example, the experiments are conducted on two high-resolution co-polarized SAR data (Capella and GF-3). The DDFormer achieves optimal detection accuracy with mean IOU (mIOU) and F-Score of 81.81% and 90%, respectively. In addition, our results are in high consistent with the Turkey Earthquake Report published by Microsoft with a correlation coefficient of 0.626. The above experiments demonstrate the robustness and effectiveness of DDFormer.

With the development of urbanisation in China, the urban areas are expanding rapidly, but there is a huge regional disparity between the east, central and western regions. The urban development in the western region lags far behind that in the eastern and central regions. In the western region of China, due to the large number of mountains and SAR backscatter mechanism, there are a lot of overlays in the image, resulting in high false alarms in built-up areas segmentation. In order to solve the problem, this paper proposed a new built-up area extraction model based on the Transformer. Different from the segmentation method based on convolutional neural network, the self-attention mechanism of the Transformer was introduced to effectively capture the image context information and reduce the impact of mountain overlays on the extraction of built-up areas. The multi-layer Transformer encoder and the multilayer perceptron (MLP) decoder were used to fuse feature maps of different scales for the sake of enhancing the ability to extract architectural features. With the purpose of improving the generalization ability, various data augmentation methods were used during training, such as random noise, random blur, and random distortion. In this paper, about 32000 samples, including some mountainous areas and around China, were used for training, which are from 27 scenes of GF-3 10m SAR images covering different areas in China. The method proposed in this paper reached a mIoU of 0.8130 and a Kappa coefficient of 0.9423, which significantly reduced false alarms in mountainous areas. Taking the study area of Lanzhou City, Gansu Province of China as an example, the result is basically consistent with the classification map of World Cover. It shows that the proposed method has a good ability to extract the distribution information of built-up areas.

In recent years, deep learning has achieved great accuracy in SAR image ship detection tasks. However, for large-scene SAR images, the sliding window method will generate a large number of slices sent to the network, resulting in a waste of GPU resources. To solve this problem, we proposed a ship detection algorithm based on CFAR, which significantly reduces the number of slices and false alarms on land. Firstly, the sea regions to be detected are obtained under the prior guidance of Global Self-consistent Hierarchical High-resolution Shorelines(GSHHS), and remove major land areas in the SAR image. Secondly, the ship candidate region generation algorithm based on block G0-CFAR is implemented to obtain the candidate slices. Finally, the Yolov5 model is used to detect ships from the candidate slices. The validation experiment of sentinel-1 and Gaofen-3 large-scene SAR images shows that the proposed method can detect the ships in large-scene SAR images with 88% detection accuracy within 97s, and the number of image slices fed into GPU is greatly reduced. It has certain practical significance for ship detection in large-scene SAR images.

In this paper, an automatic extraction method of multi-type built-up areas in SAR images is proposed. In order to adapt to the architectural differences in different regions, we improve the residual U -Net, one is to introduce multi -scale pyramid structure into the structure, the other is to use the diss loss function. The proposed method was verified using GF-3 SAR data in four regions of China and compared with FCN and PSPNET. Finally, the accuracy evaluation results show that the overall accuracy of the extraction results is greater than 86%, indicating that the method has a certain application potential.

Deep learning has been successfully applied to the ordinary image super-resolution (SR). However, since the synthetic aperture radar (SAR) images are often disturbed by multiplicative noise known as speckle and more blurry than ordinary images, there are few deep learning methods for the SAR image SR. In this paper, a deep generative adversarial network (DGAN) is proposed to reconstruct the pseudo high-resolution (HR) SAR images. First, a generator network is constructed to remove the noise of low-resolution SAR image and generate HR SAR image. Second, a discriminator network is used to differentiate between the pseudo super-resolution images and the realistic HR images. The adversarial objective function is introduced to make the pseudo HR SAR images closer to real SAR images. The experimental results show that our method can maintain the SAR image content with high-level noise suppression. The performance evaluation based on peak signal-to-noise-ratio and structural similarity index shows the superiority of the proposed method to the conventional CNN baselines.

High resolution synthetic aperture radar (SAR) data have important research significance and application prospects in crop fine classification. In order to make full use of high resolution SAR data, this paper introduced the deep learning semantics segmentation methods, which have already achieved tremendous success in the field of computer vision, into the corn fine classification using SAR images. Therefore, a corn fine classification method, based on U-Net network for high resolution SAR data, was proposed in this paper. Experiments were carried out using one scene high resolution GF-3 SAR data covering Fuyu City, Jilin Province, China. The experimental result shows that the proposed crop fine classification method based on deep learning semantics segmentation method can achieve a satisfactory classification result.

Building damage assessment with remote sensing images plays an important role in providing information for disaster rescue and reconstruction. Recently, deep convolutional networks show good ability for some remote sensing applications. However, it is difficult to obtain a large number of labeled samples for training in some cases. With regard to this problem, a transfer learning method based on deep neural networks is adopted to discriminate collapsed building from intact building in remote sensing images in this paper. Samples are obtained from high resolution remote sensing images and split into training and validation datasets. Then a collapsed building discrimination model is built based on very deep convolution networks (pretrained VGGNet via ImageNet). With the new trained VGG model, test dataset including samples of collapsed buildings and intact buildings are used to classify the two types of buildings. The preliminary experiment results show that the new trained model performed well for collapsed building discrimination.

The detection and extraction of buildings using high resolution synthetic aperture radar (SAR) images has been the topic of recent discussions. In this paper, a framework for building extraction and classification is proposed. Buildings are classified into three kinds: commercial architecture, residential building and Industrial building, by the fusion of structure features (point-like, linear and regional features). All kinds of structure features are detected orderly in the framework of multi-scales pyramid. The delineation of urban footprint is firstly determined on the highest scale level, and then the types of buildings are classified on the lowest scale level based on the local features including persistent points, corner lines, and dark shadows, which are detected in the pre-determined region of urban footprint. The experimental results on our TerraSAR-X Staring Spotlight(ST) data set confirm that our scheme performs well.

Ship classification with spaceborne high resolution synthetic aperture radar (SAR) has wide applications in maritime traffic monitoring, fishing law-enforcement operation, marine security, etc. Deep learning, which has the ability of learning features itself, is successfully used in computer vision and artificial intelligence, and introduced into remote sensing field in recent years. In this study, the Italian COSMO-SkyMed SAR images acquired on Jul. 12-15, 2010 were used for ship classification with convolution neural networks in the Google's TensorFlow environment. The results show that cargo ships could be discriminated from non-cargo ships. Due to variations of radar illumination directions and ship poses, more data are necessary for sub-category classification.

To explore the capability of quad polarimetric SAR images from a new pursuit monostatic mode of TerraSAR-X and TanDEM-X, a novel process chain for ship detection and velocity estimation is proposed in this paper. Compared with the classic processing chain used for single SAR image, more efficient techniques are integrated into this novel chain, which take advantage of the new image mode and its excellent imaging quality in term of the radiometric and geometric accuracies. Land masking based on coherence optimization and ship velocity estimation depended on the difference of positions from individual images are proposed as significant improvements in the new process chain. Their efficiency is validated with the experimental SAR images acquired over East China Sea. The experimental results show that the land regions and small islands are entirely removed and the estimation results of ship velocity is close to the ground truth from automatic identification system (AIS) data.

The new advanced very high resolution (VHR) synthetic aperture radar (SAR) sensors are capable of achieving sub-meter resolution, which offers the opportunity for a fine level of analysis of man-made structures. In this paper, we present a method for the detection and 2-D reconstruction of building radar footprints from VHR SAR scenes. The method is based on the extraction of a set of low-level features from the image and on their composition to more structured primitives using a perceptual organization method. Based on the backscattering model of gable-roofed building, building candidates are firstly detected with a circular detector proposed in this paper, then roof edges are extracted by a line detector and radar footprints are finally reconstructed. Experimental results confirm that our algorithm performs well.

Technology for marine vessel monitoring and analysis combining Synthetic Aperture Radar (SAR) and Automatic Identification System (AIS) is presented in this paper. Detection results in SAR image are validated and analyzed with corresponding AIS information. In order to evaluate the performance of the techniques, nine ERS-2 SAR images covering the year 2007 and corresponding AIS polls are used. The results were analyzed and discussed combining SAR and AIS reports. Based on the analysis, conclusions can be made that SAR and AIS have relative advantages and disadvantages respectively, using both systems together is a good way for marine traffic monitoring.

Under the condition that the accuracy of the air-borne platform is unknown, an affine transformation model made up of unknown parameters such as rotation, scale and translation is proposed for Synthetic Aperture Radar (SAR) image geometric correction. These parameters in this model have real physical meanings compared to that in the traditional polynomial models. Based on the characteristics of geometric distortions in SAR images, four unknown parameters, i.e., one rotation angle, two scale and two translation parameters, are selected for geometric correction. Different imaging look directions modes including left and right are all considered in this transformation. One real air-borne SAR image with 0.2m resolution was used to validate the availability of this method. In addition, Ground Control Points (GCP) were surveyed by the technique of Real Time Kinematic (RTK) to resolve the unknown parameters and evaluate the image geometric correction results.

In 2013, the TerraSAR-X (TSX) mission was extended by implementing two new modes, including Staring SpotLight (ST). The azimuth resolution of this mode is significantly increased to approximately 0.24 m by widening the azimuth beam steering angle range. In this letter, five types of damaged buildings due to earthquake are analyzed using TSX images in the new ST mode. In particular, the characteristics of the individual damaged buildings are investigated using the new very high-resolution images. The old Beichuan County is selected as the research area, where most of the buildings damaged by the earthquake on May 12, 2008 have been preserved. Moreover, buildings with different types of damage can be found in the study area. Features such as backscattering and texture measurements of the damaged buildings in descending and ascending post-earthquake HH and VV polarization synthetic aperture radar (SAR) images are analyzed. Visual interpretation, statistical comparison, and classification experiments are performed for the damaged building analysis. Based on the analysis, the TSX ST mode images show potential for building damage detection and for discrimination of basic damage classes. In addition, the classification results show that the gray-level co-occurrence matrix (GLCM) second moment, the variance of backscattering, and the GLCM homogeneity are the three best features for the discrimination of damage type.

A method to estimate absolute phase of the target, and phase compensation for the coherence matrix are introduced in this paper. The absolute phase of a target relates with the generator of target local shape, and with the the attenuation coefficients of vertical and horizontal polarimetric wave. The effectiveness of phase compensation is validated via Yamaguchi decomposition with phase compensation of Radarsat-2 full polarimetric data collected over Suzhou City, China. The estimation of absolute phase of target and the phase compensation are validated and benefit to correctly interpret the scattering mechanisms of targets, with decreasing the number of negative power pixels.

High-resolution synthetic aperture radar (SAR) data have been widely used in marine environmental protection, marine environmental monitoring, and marine traffic management. Ship detection is one of the important parts of SAR data for marine applications. This letter focuses on the feature analysis of ships in high-resolution SAR images and proposes an improved optimizing algorithm for ship detection. A fast block detector is designed to extract sea clutter in a uniform local area, and then a constant false alarm rate detector is employed. Based on the kernel density estimation of ships, aspect ratio, and pixel points, ships are identified. TerraSAR-X and COSMO-SkyMed images are used to test our algorithm. The experimental results show that this algorithm can be implemented with time-saving, high-precision ship extraction, feature analysis, and detection.

Ship monitoring has a wide range of applications in marine activities and maritime management. Spaceborne synthetic aperture radar (SAR) technology has an advantage in ship detection over a vast region compared to other technologies. Recently, high-resolution SAR satellites have made ship recognition and classification possible from space. In July 2010, a ship recognition campaign was performed in the East China Sea. In situ investigation of ship types was carried out during the periods when the Italian COSMO-SkyMed satellites passed over the test site. Automatic Identification System (AIS) information was collected. Based on this information, a novel hierarchical ship classifier for COSMO-SkyMed SAR data was proposed. A total of 41 ship chips were cut from the SAR images for later classification. After preprocessing of the ship chips, geometric and backscattering characteristics of various ship types were analyzed. The ships were classified into bulk carriers, container ships, and oil tankers, with an accuracy of 93.3%, 80.0%, and 72.7%, respectively. Further investigation of the backscattering features with various illumination conditions is still in progress.

Recently, there has been an increasing interest in compact polarimetry (CP), which helps to reduce the complexity, cost, mass, and data rate of synthetic aperture radar (SAR) systems while attempting to maintain many capabilities of a fully polarimetric system. In this letter, we provide a comparison of the capabilities of three conventional CP modes to distinguish different physical scattering mechanisms (PSMs) through H-α decomposition. A compact H-α space for dual circular polarimetric mode is proposed on the basis of the distribution centers and densities of different PSMs. Both airborne and spaceborne SAR data are used to validate our method. The experimental results show that the decomposition accuracy of each PSM is improved compared with the results of previous publication, the kappa coefficient increases by nearly 0.1, and the overall accuracy is above 81%.

In this paper, we present a novel unsupervised change detection scheme for multilook polarimetric synthetic aperture radar (PolSAR) images using heterogeneous clutter models. First, a multilook product model is introduced to describe the heterogeneous clutter for multilook PolSAR data, and a corresponding covariance matrix estimation method is derived. Based on this model, a new similarity measure is proposed to quantify the degree of evolution between the statistical characteristics of multitemporal fully PolSAR images. Compared with the classical similarity measure of Wishart distribution, both the structure of the covariance matrix and the power information of PolSAR clutter are considered in the proposed similarity measure. A Kittler and Illingworth (K&I) minimum-error threshold segmentation method is applied to extract the changed areas. Both the simulated PolSAR data set and two three-look Radarsat-2 fully polarimetric images of Suzhou, China, acquired on April 9, 2009 and June 15, 2010, are used for our experiment. The results demonstrate that the proposed change detection method can give a much higher detection rate and a lower false alarm rate than the method using the Wishart similarity measure.

Ship wake detection can facilitate ship detection and ship motion parameters retrieval. In this paper, we propose a novel constant false alarm rate (CFAR) ship wake detection algorithm in SAR images based on length normalized scan method. As we all know, range component of ship's movement leads to azimuth offset in SAR images. The start point of wake should be located near the ship in the azimuth. According to this physical facts, length normalized scan is performed on line which may be a wake. By means of linear integral and length normalization technology, linear feature detection is transformed to point detection. Then the probability model in the scan domain is constructed to implement CFAR detection. Different resolution real SAR images are used to validate our proposed method. Experimental results show that this algorithm is effective for ship wake detection and can extract ship's velocity with good accuracy.

In this paper, we propose a polarimetric TomoSAR method with Wiener-SVD inversion and Gerschgorin Disks for different scattering mechanisms of each point. Seven images of RADARSAT-2 Fine Quad Mode have been used to validate the efficiency of the proposed method. The polarimetric analysis shows that more scattering points can be detected by the composition of different polarimetric bands, thus retrieving more information on the elevation within one resolution cell.

Synthetic Aperture Radar (SAR) is a significant tool to satisfy the growing demand of the maritime vessel traffic. ScanSAR, as an important role of SAR systems with very high imageries swath, is more suitable for the detection issue. In this paper, features of ships on Radarsat-2 ScanSAR imagery are characterized as “Bright-Dark” structure. According to the unique features, a new ship detector based on the Difference of Gauss (DoG) scale-space is proposed. To enhance the robustness, a threshold method simulated by the “dark spots” matrix is designed. In the threshold method, the average KL test is carried out with 6 different distributions on several Radarsat-2 ScanSAR Narrow imageries of sea which shows the Gamma distribution fits the sea clutter the best and is selected. Finally, the proposed detector is validated on a slice of Radarsat-2 ScanSAR imagery and comparisons with CFAR are made.

In this paper, vessel characterization is investigated using fully polarimetric Radarsat-2 SAR data. Polarimetric decompositions, including coherent decompositions, decomposition based on eigenvector analysis and model based decomposition, are applied to the quad-pol data to investigate vessel's scattering mechanism. Corresponding AIS and vessel photographs were acquired as ground truth data to facilitate analysis. Three types of vessels, such as bulk carrier, oil tanker and container ship, were investigated. Results show that the stable dominant scatters in a vessel and their scattering mechanism are expected as efficient features for vessel classification.

This paper introduces a new compact polarimetric (CP) three-component decomposition scheme, which can overcome the overestimation of volume scattering power in urban areas to a certain extent. For CP three-component decomposition, degree of polarization (DoP) is one of the key parameters, and directly influences the effectiveness of the decomposition method proposed in [1]. To overcome the overestimation of volume scattering component, a new method for estimating the DoP is introduced, and with the new DoP, volume scattering power decreases significantly. By considering the original decomposition method performed effectively in forest, lawn and ocean, the new DOP is only applied in urban areas. CP data in CTLR mode simulated from Radarsat-2 C-band fully polarimetric (FP) image is used to demonstrate the validity of the new decomposition scheme.

The high-resolution images, acquired by the new SAR sensors, provide opportunities to detect changes more accurately. The traditional pixel-based change detection method cannot adapt to the change detection of objects in different spatial scales, especially in build-up area, which has strong speckle noise. Recently, Object-based image analysis provides a feasible idea to solve this. In this paper, a SAR change detection method based on object-based image analysis is proposed. In proposed method, temporally sequential images are combined and segmented together to produce spatially corresponding image-objects. With different parameter settings, the segmentation maps of different scales are generated, which result in different difference maps. To utilize the information of different scales, a fusion method is employed to generate the final difference map, and a binary change detection map is produced on the basis of the difference map. The experiments show that object-based method is more efficient than the pixel-based method.

This paper proposes a new detector for interested object detection in high resolution SAR image. To obtain higher performance, four parameters are required to describe features of the interested object. Specifically, the parameters of segmentation scale and average expected object size are used for SAR image segmentation to get patches, which are small image parts to capture fine scale information such as gray variance and shape. In this step, adjacent map is adopted to recode the information and the relationship of all the patches; The third parameter is false alarm rate, which is used to retrieve the intensity threshold for finding out all the potential targets in SAR image; The parameter of object packing density is finally used to reduce the false alarm, which is calculated according to their occurrence frequency. To evaluate the proposed detection methodology, air-born SAR image with one-meter resolution is tested. The scene context of this image is complex, since it includes building, road, vegetation and trucks. If the trucks are taken as the interested objects, this methodology can detect all of them with lower false alarm rate.

Ship classification in high-resolution synthetic aperture radar (SAR) satellite images is a hotspot and a continuing problem in SAR applications. The scattering components of ships are the strong scatter of objects in SAR images, and these can represent the superstructure of different ship types. Based on analyses of different scattering components of bulk carriers, oil tankers, and container ships, we propose a new classification method for these three ship types in COSMO-SkyMed SAR images. First, morphological preprocessing is applied to suppress sidelobes. Second, based on Hough transform (HT), the orientation of the principal axis is extracted, and the modified minimum enclosing rectangle (MER) of the ship is obtained and rotated along the principal axis. Finally, the ship type is decided according to the width ratio of MER between the HT line, the ratio of ship and nonship points on the principal axis, and the scattering density. The results show that this method has good performance in ship classification, with an overall accuracy of over 80%.

Ship detection and monitoring techniques are important applications in protection of marine resources. In this paper, a ship detection method is proposed based on the Freeman three component decomposition and OS-CFAR detector. The detector has been applied on RADARSAT-2 QUAD Mode image to demonstrate the efficiency of the method. Compared with polarimetric white filter, SPAN image, HH-CFAR and HV-CFAR methods, the proposed method can detect ships in ocean clutter correctly and control false alarms well. Moreover, this method does not depend on the size of targets and therefore offers more adaptability for ship detection.

Polarimetric information can be used to characterize the target and benefit for ship classification in SAR image. In this paper, three types of features from fine quad-polarization Radarsat-2 SAR image, such as target to clutter ratio, distribution of scatter point, coherent decomposition component, are analyzed of three types of ships such as bulk carrier, container ship and oil tanker. Preliminary results show that target to clutter ratio is not suitable to discriminate the three types of ships. HH polarization is optimal for ship detection to HV,VH and VV polarization under the condition of the incidence angle of about 48 degrees. Distribution of strong backscattering can reflect the geometry structure of ship in SAR image, and it can be a good feature for ship classification. Furthermore, the feature will be more reliable in higher resolution SAR image. Polarimetric coherent decomposition benefits for understanding scatter mechanism of scatters of different ship. However it has some limitations for ship type recognition. More features and more ship chips should be investigated for the further conclusion.

It's worth to investigate that the false alarm is reduced with polarimetic information while keeping the detection ratio when the fully polarimetic SAR images used for ship detection. An improved S-NMF (Sparseness-Nonnegative Matrix Factorization) for ship detection according to the nonnegative and sparseness characteristics of eigenvalues is proposed in this paper. The first two eigenvalues are used for NMF decomposition since they take most of the energy of the targets, and then we combine the advantage of the third eigenvalue which could get rid of the false alarms for ship detection with the OS-CFAR method. The proposed method is validated using Radarsat-2 fully polarimetric images with the corresponding AIS data. Results show that it cannot only detect most of the ships, but also reduce the false alarms effectively, and has good ability when the detection conditions change.

This study presents a dynamic deformation retrieval method for differential synthetic aperture radar interferometry (DInSAR) processing. Detailed nonlinear displacement parameters are modeled and estimated for capturing the developing trends of ground movement both in time and space. In order to reduce the system disturbance errors such as atmospheres and platform deviations, the singular value decomposition technique is applied. 13 ascending ALOS PALSAR images from 2008/07/05 to 2010/08/26 are collected for the algorithm experiment. The studied area locates in Tianjin China and detailed deformation evaluation patterns have been obtained. The experimental results are validated via leveling data and prove the effectiveness of the algorithm.

Vessel detection becomes more and more important applications of remote sensing. In this paper, a vessel detection method based on gravity enhancement and multi-scale processing was applied to Synthetic Aperture Radar (SAR) images. With the method, targets' pixels were enhanced by the interactions between themselves and their neighbors, while the speckle and background clutter were suppressed after enhancement as well as the contrasts between targets and clutter had been greatly increased. In high resolution SAR image, vessel includes more pixels and can show its structure in the image. Traditional Constant False Alarm Rate (CFAR) detector which is proper for low resolution SAR image may lead to more false alarms if it is applied to high resolution image. So before segmentation, multi-scale processing was applied to avoid this problem. Cosmo-Skymed image with 3 meter pixel spacing was adopted to test the algorithm. Results show good ability for vessel detection with the high resolution SAR image.

In order to expand SAR application into vehicles monitoring, it is necessary to analysis and model the vehicle's characteristics under different SAR imaging conditions. In this paper, three typical trucks at different aspect angles were imaged by Radarsat-2 quad-polarimetric SAR image with different incidence angles in the field. RCS measurement and polarimetric decomposition were used to analysis the variation of vehicles phenomenon in SAR image. Some benefited conclusions are drawn based on analysis of the curve of RCS and polarization component which are varying with the aspect angle, incidence angle and structure.

This letter presents preliminary results of an attempt to monitor rice crop growth using RADARSAT-2 quad polarization synthetic aperture radar (SAR) data. Three RADARSAT-2 quad-polarization SAR images are collected from transplanting to rice crop harvesting. Ground truth data, such as rice height and biomass, are measured during RADARSAT-2 data acquisition in Hainan Province, South China. The correlation between backscattering coefficient and rice growth parameters is analyzed, and then, a rice field mapping method with quad polarization SAR image is developed. Experiments show that an HV or VH image backscattering coefficient exhibits the best correlation with rice age after transplantation. Furthermore, the HV or VH image is also more suitable for retrieving rice growth parameters, such as rice height and dry biomass, for FQ4 RADARSAT-2 SAR data. The ratio image of HH/VV possesses high separability required to distinguish rice crop from banana, forest, and river. Results indicate that RADARSAT-2 quad polarization SAR data presented enormous potential for monitoring rice crop growth.

This paper presents a SAR image change detection algorithm for urban areas which mainly contains three steps: (i) modeling the Synthetic Aperture Radar (SAR) image by G0 distribution; (ii) generating a change map by computing the Kullback-Leibler divergence; (iii) masking the change map to obtain the final change areas. We propose to use the second kind statistics based parameter estimation method via Mellin transform to figure out the three parameters of G0 distribution. Experiment results indicate the second kind based G0 distribution model outperforms the moment based methods and can achieve a satisfactory result.

Radarsat-2 satellite offers general users the Quad-Pol SAR image service with a resolution of 8 meters, which provides valuable data source for the research of traffic vehicles monitoring on Quad-Pol SAR images. According to the vehicle target in such new SAR images, this paper put forward a target characteristic analysis method, which is a combination of target RCS measurement and polarization decomposition. Moreover, it choose the large trucks as an example, give out a conclusion of characteristic analysis of vehicle targets through the on-site synchronous experimental data obtained at different incident angle. Thus provides the research foundation for further exploration of realizing the traffic monitoring with space-born Quad-Pol SAR images.

On May 12, 2008, Wenchuan earthquake (Ms 8.0) occurred in Sichuan, Southwestern China. This catastrophe caused severe damage of constructions in urban and rural areas, and fundamental infrastructure, such as those facilities of factories, electrical power, telecommunication and transportation, etc. The earthquake also resulted in many geological phenomena, i.e. landslide, debris flow, landslide lakes, etc. which also triggered off damage and threat. The airborne campaigns were performed after the earthquake with high-resolution SAR system for rescue and relief effort and damage assessment. In this study, damage of various facilities was described and analyzed using airborne SAR images acquired during the campaigns. The results were verified with ground truth investigation. The results showed the role of SAR data in earthquake damage assessment.

A new method is proposed to get the segmentation threshold and detect the dark spot in oil-spill images. The method is inspired from the K -distribution model of sea background, which is widely accepted to describe the ocean clutter. By comparing the histograms of oil-spill region and the sea background, it is found that the oil slicks break the K -distribution model, but there is still some information unchanged-the relative probability ratios among the pixel values in 95%~99% CDF extent, which is used to deduce the original κ -distribution model. Finally, the intersection of the original histogram of the oil spill image and the deduced sea background PDF is selected to be the threshold. Experiment in RADARSAT-2 image shows the effectiveness of the method.

In this paper, change detection with multi-polarization SAR imagery is focused. Based on the polarimetric statistic distribution, a polarimetric test statistic is applied to evaluate the equality of two areas in two pass polarimetric SAR images. And then, in order to find out the `real' changed area, a constant false alarms rate (Pfa) is given to determine a threshold for change detection. Moreover, a majority processing considering the context information of a given target is preferred to improving the final accuracy of the change detection. Finally, the proposed method was tested with the multitemporal Envisat-ASAR's alternative polarization SAR data. Experiments are performed to validate efficiency of the method presented in the paper.

In this paper, we investigate a case-based reasoning (CBR) method for the classification of multi-temporal SAR images with the aid of ancillary information. Our scheme for the problem of multi-temporal SAR images classification comprises four main steps, including SAR image processing, construction of case library, case-based classification and post- classification processing. During the construction of case library, we employ a spatial-temporal analysis technique to remove fake cases, which can guarantee cases with high confidence. In the implementation of case-based classification, we propose a similarity assessment and use it for the case-based matching. After that, we investigate an object-oriented post-classification method which takes the shape of land use region into account, as a result, it leads to a more meaningful classification, and the regenerate land use image or map can be easier compared and combined with usual GIS data. Multi-temporal ENVISAT ASAR images from 2004 to 2005 are used in our experiments, where their resolution are 12.5 times 12.5 m. The study site is located in Beijing, China. During our experiments, we use the land use map of 2004 to assist the construction of the case library. The results of our experiments indicate that the CBR method is very promising for the classification of multi-temporal SAR images, where the overall classification accuracy can reach up to 80%.

A wavelet based automatic targets detection method for high resolution airborne SAR data is described in this article to receive faster and more accuracy detection. This method is based on the assumption that man-made objects are easily detectable at low resolution because their scattering is more persistent than that of natural objects. The algorithm involves an improved wavelet soft threshold filter (IWSTF) and a wavelet based RCCFAR detector. In order to retain the target feature, the wavelet soft threshold filter is improved by the strategy used in the enhanced Lee filter. Instead of using a global threshold, we adopted an adaptive threshold calculated according to the detail coefficients in each scale. To accelerate the RCCFAR detector, two RCCFAR detectors are used. One is first applied to the approximate coefficients to make a coarse detection. The other one is applied to the filtered images in those regions which are regarded as candidate targets. Performance of the algorithm is assessed by some high resolution airborne SAR image and it shows that the algorithm can effectively reduce false alarms caused by speckles.

In the present paper, according to data analysis, the problem of change detection has been addressed with unsupervised change detection with multi-temporal single- channel single-polarization RADARSAT-1 SAR images. Method proposed in this paper is based on four main steps: (1) Data preprocessing. SAR images are rectified and calibrated to Gauss-Kruger projection firstly. Moreover they are co-registered to ensure the condition of good accuracy of change detection. (2) Change detection. The window log-ratio approach is adopted to compare multi-temporal SAR data, because of the multiplicative speckle in the images. Then a proper window is selected to average pixels from the two SAR data in log-ratio process instead of pixel by pixel. (3) Selection of optimal threshold value. Minimum error threshold is applied to find the optimal threshold. (4) Change detection map can be obtained by segmentation of window log-ratio image. Experiments show the approach deal well with change detection in water area.

Residential area extraction from remote sensing data plays an important role for map updating. In this paper, a fused method is proposed to extract residential area information originating from China airborne SAR and optical data. First, the airborne SAR and optical images are registered. Then texture analysis, using gray level co-occurrence matrix (GLCM) texture features, is applied to the SAR image. According to those texture features, with a proper threshold, the silhouette of building area can be obtained. With the optical multispectral image, vegetation plots in residential area are extracted. The logic OR is applied to fuse the processed results of the SAR and optical images. Finally, the coverage of residential areas is overlaid on the optical image to check the effectiveness of this method.

Unlike optical images, the microwave radiometer image contains information of ground brightness temperature. To integrate some features of microwave radiometer images and optical images, some transform schemes, such as intensity-hue-saturation (IHS), principal component analysis (PCA) and wavelet transform (WT), were applied to merge the airborne synthetic aperture radiometer images and Landsat ETM+ images. The experiments show that the wavelet based transform method shows the best potential among the three methods.

Image correction and preliminary analysis of a field measurement by a C band 6-channel airborne microwave synthetic aperture radiometer is presented. This measurement was carried out along the Yellow River banks, Shaan'Xi, China, during 25/sup th/ to 30/sup th/ April, 2002. The CSSAR one dimensional C-band (6.6 GHz) 6-channel synthetic aperture radiometer was used as the airborne sensor. On board of the helicopter, a GPS receiver is also used to record the geographical positions of the flight pass. The scan was carried in over a square area across the experiment fields covering about 8/spl times/8 km/sup 2/. Ground truth observation also took place at the same time. The ground measurement parameters include: temperature of water body, air temperature, soil temperature, soil moisture at varied depth, soil surface roughness, etc. Geometric correction and registration were done before the analysis of the imagery. Some preliminary results are given.