Most researches on indoor target trajectory tracking using millimeter-wave radar have faced challenges such as interference from multipath effects, which led to corrupted point clouds, and large tracking errors in multi-target scenarios. This study aims to improve the accuracy of human multiple-object tracking, addressing the practical challenges of target trajectory tracking in indoor environments. We proposed an extended Distributed Radar Multi-Target Tracking (DRMTrack) framework that enabled the fusion of point clouds from multiple radar nodes. Additionally, by exploiting the spatial distribution characteristics of the target point clouds for clustering, tracking filtering and integrating historical data for tracking association, the framework enhances multiple-object trajectory tracking performance. Experimental results demonstrate that for various trajectory paths, the minimum position tracking error for multi-target is 5.2 cm. In the five-target tracking scenario, the 90th percentile tracking error is 18.8 cm, representing a 27.7% improvement in accuracy compared to single-radar tracking. The DRMTrack system effectively reduces interference from clutter point clouds, enhances multiple-object tracking precision, and supports real-time computation. This system is suitable for motion monitoring in indoor environments such as homes and hospital rooms, effectively fulfilling the needs of real-time, non-contact health management.

The multi-agent consensus equilibrium (MACE) mechanism possesses the notable capacity to incorporate multiple priors for SAR imaging and feature enhancement. We integrate the MACE mechanism with the combined dictionary (CD) regularization model and establish an optimized multi-feature sparse imaging method. The optimized method can effectively reduce storage and computational cost. While solving the parameter selection criteria of the proposed method, real data from a Ku-band SAR amounted on a Unmanned Aerial Vehicle is utilized to evaluate the performance.

With the expansion of the surveillance area and resolution of spaceborne synthetic aperture radar (SAR) systems, the increasing amount of echo data requires further research on efficient imaging methods. The pulse repetition frequency (PRF) of traditional SAR needs to satisfy the Shannon–Nyquist sampling theory, while the PRF limits the system swath width, making it impossible to achieve wider illumination coverage. The reduction of the PRF can effectively increase the swath width, but it will cause severe azimuth ambiguity, decreasing the image quality. Several methods have been introduced for ambiguity suppression, but many of them become ineffective at lower PRFs. In this article, we propose a novel sparse imaging method for the spaceborne PRF-reduced sliding spotlight SAR. The proposed method considers the azimuth ambiguity term in the joint sparse imaging model, separately constrains the main imaging and azimuth ambiguity areas, and achieves azimuth ambiguity suppression by using compressive sensing (CS) technology. With its help, we can reduce the original PRF by up to half to double the swath width and enable high-precision sparse reconstruction of large-scale scenes. Compared with$L_{2,1}$-norm regularization-based algorithm, the proposed method shows the superior ambiguity suppression ability with less computational cost from PRF-reduced echo data. Experimental results on simulated raw data validate the proposed method.

Most existing deep learning models for remote sensing scene classification (RSSC) adopt offline learning paradigm, which are trained on closed datasets and fail to dynamically update with new class data. Currently, class-incremental learning (CIL) allows models to learn new classes while retaining discrimination of old ones. However, most CIL approaches aim to overcome catastrophic forgetting by employing techniques such as exemplarmemory and knowledge distillation, while ignoring the prediction bias caused by imbalanced datasets, where old classes retain fewer samples than new ones. Moreover, they do not adequately account for the multilevel semantic structure and multiscale feature information inherent in remote sensing images (RSIs). To address these issues, we propose an effective CIL framework for RSSC, named class bias correction network (CBCNet). Specifically, a cross-dimensional and interaction-aware attention mechanism (CIAM) is designed to incorporate channel, position, and direction-aware information in feature maps, enabling the model to highlight informative regions within RSIs. Next, a contextual information fusion module (CIFM) is proposed to explore the correlations among multilevel features and enhance representation quality through their fusion. In addition, the designed taskwise classifier head decoupling mechanism (TCDM) imposes a constraint to mitigate the prediction bias toward new classes, and enhances model’s discrimination among all seen classes. Finally, a multilevel integrated knowledge distillation module (MKDM) is developed to ensure comprehensive knowledge transfer, empowering the model to maintain critical representations in feature space and make well-informed decisions in output probability space. Experiments on five open datasets demonstrate the outperformance and robustness of our method.

Airborne synthetic aperture radar (SAR) systems are commonly susceptible to trajectory deviations, resulting in distinct azimuth phase error in the collected echo. SAR autofocus technology can compensate phase error and produce a well-focused image based on echo data. However, due to the influence of unfavorable factors such as radar interruption and electromagnetic interference, the echo data may be down-sampled in azimuth, which reduces the phase error estimation accuracy of traditional autofocus methods. By introducing compressed sensing (CS) to SAR data processing, sparse SAR imaging shows outstanding performance in acquiring high-resolution images utilizing down-sampled echo. However, the phase error existing in azimuth direction will reduce the sparsity of observed scene, and the precision of sparse reconstruction is consequently decreased. This paper aims at enhancing the Doppler rate error estimation precision when echo is down-sampled in azimuth and proposes a novel sparse imaging method combined with Doppler rate estimation. During each iteration of complex approximated message passing (CAMP) algorithm, the Doppler rate error is estimated according to the non-sparse solution by fractional Fourier transform (FrFT). Then phase error is compensated to the non-sparse solution, and the azimuth matched filtering (MF) operator is upgraded. The aforementioned steps are performed iteratively until a well-focused sparse SAR image is generated. It should be noted that the Armijo rule and random sample consensus algorithm (RANSAC) are introduced to guarantee the fast and precise reconstruction of the observed scene. Experiments from simulated and airborne data prove the enhancement in Doppler rate estimation precision by the proposed method than traditional estimator when used data is azimuth down-sampled.

In traditional office management, it is labor-intensive to perform real-time oversight on equipment and personnel. To address this challenge, this paper proposes a digital twin-based office management system. The system leverages the ESP8266 wireless module for device control and data collection, and employs the YOLOv5 deep learning model for real-time detection of employees’ working conditions. Additionally, a virtual office environment is constructed using the Unity engine. The system implemented herein enables real-time monitoring and analysis of office utilization, and assists managers in optimally allocating resources to enhance resource utilization efficiency by leveraging intelligent sensing and decision-making technologies. The system incurs low hardware and software costs, minimal data transmission latency, and rapid response times across its modules. Moreover, through data masking techniques, the system can protect the privacy of office personnel while enabling real-time monitoring.

Video Salient Object Detection (VSOD) is a significant pre-work for many vision applications. Different for Salient Object Detection (SOD), an effective VSOD model requires not only the spatial domain of origin image but also temporal domain. In this paper, we proposed an optical flow guided pyramid network (OFPN) for VSOD, which exploit the temporal optical flow (OF) to assist VSOD. Due to the fact that optical flow maps have slightly lower quality compared to depth maps, we designed two modules for seeking better improvement. To this end, we render optical flow maps from RGB images firstly. Then, an adaptive cross-modal attention module (ACA) is designed for multi-modal fusion. The high-level encoded features are aggregated into a shared decoder for primary prediction. Besides, the low-level features are separately sent into multi-scale context attention module (MCA) for multi-scale context fusion with the assist of the primary prediction level by level. Further, we exploit a multi-scale loss to take full advantage of the hierarchical details through image pyramid structure. Extensive experiments on five benchmark datasets demonstrate the superiority of our method against 12 state-of-the-art methods.

This paper presents a Unity engine-based digital twin intelligent decision-making system for campuses, aiming to improve campus management efficiency and student experience. The system combines shapefile information and tilt-shot fusion technology to achieve 3D reconstruction of campus buildings and environments, creating a digital twin model. For intelligent decision-making, we simulated a virtual energy environment and applied a reinforcement learning algorithm to address real-world energy decision challenges. Concurrently, IoT technology is used to monitor campus devices and resources, including energy utilization, security, and environmental quality. The integrated data is presented through a visual and interactive interface for real-time monitoring and management of campus resources by administrators and students. This system enhances resource utilization efficiency, sustainability, and security, showcasing the practical application of digital twin technology in modernizing campus management and improving the student experience in institutions.

Azimuth ambiguity in synthetic aperture radar (SAR) images results from the limited azimuth sampling rate and seriously affects image quality and normal applications. Multiple algorithms for azimuth ambiguity elimination have been proposed. However, most algorithms adopt a single-channel model and ignore the effect of additional reconstruction operations on ambiguity in azimuth multichannel systems (MCSs). In this letter, the particularity of azimuth ambiguity for MCSs is analyzed theoretically. A refined ambiguity suppression scheme based on multichannel cancellation (MCC) and ambiguity refocusing is proposed. The ambiguity region is first located through the comparison of multi-look image pairs. Then the interference image for ambiguity extraction is cancellated through MCC. Consequently, the ambiguous image is refocused and extracted. Finally, the ambiguity is reversed to the echo domain to handle the special replicated ambiguity problem for MCSs. With the proposed scheme, ambiguity suppression performance has improved for MCSs. The validity of the proposed algorithm is further verified through LuTan-1 real data.

Vision-language models have demonstrated impressive capabilities in associating images and text by pretraining on extensive image-text paired data. The paradigm of continual pretraining followed by fine-tuning has become prevailing for boosting performance in domain-specific tasks under constrained computation resources. Benefiting from the superior generalization abilities of foundation models, the demands for computational resources and extensive data corpora have been significantly reduced. Nonetheless, it is crucial to tailor the model for the characteristics of downstream tasks to mitigate the misalignment between the pretraining pretext tasks and actual applications of interest. In this study, we utilize a CLIP-based model that has been continually pretrained on the 5 million image-text dataset in the remote sensing field as the foundation model, focusing on cross-modal image-text retrieval tasks. We introduce an efficient framework called remote sensing image-text retrieval fine-grained fine-tuning (RSITR-FFT), which refines the feature space by introducing fine-grained word-region alignment and incorporating consistency constraint regularization terms in the learning objectives. The fine-grained alignment aims for precise word-region correspondence beyond classical global-level image-text matching, while the consistency regularization encourages geometric coherence between the image and text modalities. Remarkably, our method achieves observable performance improvements while requiring far fewer fine-tuning samples—about 10 000, in contrast to the 400 million and 5 million samples used during the CLIP’s initial pretraining and GeoRSCLIP’s continual pretraining stages, respectively. We perform quantitative evaluation on RSICD, NWPU-Captions, and UCM-Captions datasets to demonstrate the effectiveness of RSITR-FFT. We further showcase its realistic application on the high-resolution remote sensing imagery through the qualitative visualization experiments on the FAIR1M-1.0 dataset. The code and models are available athttps://github.com/d1x1u/RSITR-FFT.

The azimuth multichannel synthetic aperture radar (SAR), as one of the mainstream technologies for achieving high-resolution and wide-swath (HRWS) imaging, has been successfully employed in several on-orbit SAR missions. However, the unavoidable phase errors among channels result in azimuth ambiguity, deteriorating the recognizability of targets in SAR images. Additionally, the radio frequency interference (RFI) exacerbates the difficulty of phase error estimation. To address this issue, a phase error calibration method based on least spectrum difference (LSD) is proposed. Firstly, the multichannel signals are reconstructed using the linear mapping form of the reconstruction algorithm. Secondly, the objective function is established based on the continuity of the azimuth spectrum, by which only the signals near the discontinuity points are proposed. Finally, the optimal estimation of the phase differences can be obtained after iteration. In LSD method, the RFI to the objective function is mitigated due to the operation in the range-Doppler domain, and the iteration is proposed only using a few data near the discontinuity points thus saving much computational cost. Experimental results based on the simulated data and real bistatic echoes of the LuTan-1 (LT-1) mission validate the superiority of the proposed LSD method.

Range ambiguity is a technical challenge due to the deterioration of the image quality in the multistatic synthetic aperture radar (SAR) constellations. This article proposes an effective range ambiguity suppression scheme based on multiechoes coherent processing. First, a general signal model impacted by range ambiguity is built up based on the geometry of the multistatic SAR constellation, showing the different phase characteristics of the desired and ambiguous signals between the satellites. Then, an effective processing scheme for range ambiguity suppression is proposed based on the different phase characteristics, and the corresponding ambiguity suppression performance (ASP) is analyzed accordingly. This scheme can achieve a coherent summation of the desired signals by compensating for the corresponding phase difference, while the ambiguous signals are summed incoherently. Therefore, the range ambiguities can be suppressed without increasing the SAR instrument complexity. Finally, the system and imaging simulation results are provided to validate the theoretical analysis of the ASP and demonstrate the effectiveness of the proposed scheme, respectively.

The multiagent consensus equilibrium (MACE) mechanism, which generalizes the popular method plug-and-play (PnP)-alternating direction method of multipliers (ADMM) and composite regularization in computational sensing, possesses the notable capacity to incorporate multiple priors aspect of both regularization and PnP for improving image quality. In this work, a flexible synthetic aperture radar (SAR) image reconstruction method based on MACE is proposed to integrate multiple regularization and PnP priors for various features enhancement. The partial-update approach and Mann iteration methods are implemented to increase the computational efficiency of the MACE-based SAR image reconstruction algorithm. A thorough analysis of the proposed algorithm’s convergence and computational complexity is provided. High-quality SAR images necessitate low ambiguity, high target-to-background ratio (TBR), and low coherent speckle. We therefore demonstratively integrate regularization and PnP priors for azimuth ambiguity suppression, sparsity inducing, multiple features enhancement, and despeckling. The proposed method’s performance is evaluated through experiments on both simulated and QILU-1 satellite SAR data.

Synthetic aperture radar (SAR) tomography (TomoSAR) retrieves three-dimensional (3-D) information from multiple SAR images, effectively addresses the layover problem, and has become pivotal in urban mapping. Unmanned aerial vehicle (UAV) has gained popularity as a TomoSAR platform, offering distinct advantages such as the ability to achieve 3-D imaging in a single flight, cost-effectiveness, rapid deployment, and flexible trajectory planning. The evolution of compressed sensing (CS) has led to the widespread adoption of sparse reconstruction techniques in TomoSAR signal processing, with a focus on ℓ1norm regularization and other grid-based CS methods. However, the discretization of illuminated scene along elevation introduces modeling errors, resulting in reduced reconstruction accuracy, known as the “off-grid" effect. Recent advancements have introduced gridless CS algorithms to mitigate this issue. This paper presents an innovative gridless 3-D imaging framework tailored for UAV-borne TomoSAR. Capitalizing on the pulse repetition frequency (PRF) redundancy inherent in slow UAV platforms, a multiple measurement vectors (MMV) model is constructed to enhance noise immunity without compromising azimuth-range resolution. Given the sparsely placed array elements due to mounting platform constraints, an atomic norm soft thresholding algorithm is proposed for partially observed MMV, offering gridless reconstruction capability and super-resolution. An efficient alternative optimization algorithm is also employed to enhance computational efficiency. Validation of the proposed framework is achieved through computer simulations and flight experiments, affirming its efficacy in UAV-borne TomoSAR applications.

Remote sensing technology has made remarkable progress, providing a wealth of data for various applications, such as ecological conservation and urban planning. However, the meticulous annotation of this data is labor-intensive, leading to a shortage of labeled data, particularly in tasks like semantic segmentation. Semi-supervised methods, combining consistency regularization with self-training, offer a solution to efficiently utilize labeled and unlabeled data. However, these methods encounter challenges due to imbalanced data ratios. To tackle these challenges, we introduce a self-training approach namedDAST(Dynamic andAdaptiveSelf-Training), which is combined with dynamic pseudo-label sampling, distribution matching, and adaptive threshold updating. Dynamic pseudo-label sampling is tailored to address the issue of class distribution imbalance by giving priority to classes with fewer samples. Meanwhile, distribution matching and adaptive threshold updating aim to reduce distribution disparities by adjusting model predictions across augmented images within the framework of consistency regularization, ensuring they align with the actual data distribution. Experiment results on the Potsdam and iSAID datasets demonstrate that DAST effectively balances class distribution, aligns model predictions with data distribution, and stabilizes pseudo-labels, leading to state-of-the-art performance on both datasets. These findings highlight the potential of DAST in overcoming the challenges associated with significant disparities in labeled-to-unlabeled data ratios.

The current synthetic aperture radar (SAR) images with ultrahigh-resolution provide detailed structures of the urban areas. Utilizing stacks of ultrahigh-resolution SAR images acquired with different view angles, tomographic SAR (TomoSAR) becomes an advanced technique to retrieve 3-D spatial information of the detailed structures which presents the efficient density in the point clouds. The technique is a sparse reconstruction problem indeed and can be solved by compressive sensing (CS) algorithms. However, conventional CS algorithms process the pixel independently and the detailed structures of the targets are easily lost followed by the sparsity constraints. In this article, we apply morphology regularization as a prior term to form a novel approach based on the CS algorithm. The morphology regularization enhances the detailed 3-D structural properties of targets, which can shrink the concatenations caused by outliers in the iterative reconstruction. As for the optimization algorithm for tomographic inversion, we apply the framework of the alternating direction method of multipliers (ADMMs), where the Bregman iteration is adopted for solving the subproblem with morphology regularization. Both simulation experiments and tests on real data show that the proposed method can suppress the outliers and guarantee the detection rate, leading to excellent correctness and completeness.

There is an urgent need for industrial manufacturing to fully integrate with emerging technologies to build enterprise core competitiveness. Currently, existing methods have difficulty meeting the high-precision and stability practical requirements with diversified industrial products. In this study, a PMAE (Parallel Multiple AutoEncoders, PMAE) model is proposed, which is designed with parallel multiple encoders based on the AutoEncoder framework. It uses the network of parallel multiple encoders with different encoder structures to obtain latent features that have rich and precise semantic information. A consistency objective function is proposed to make the PMAE network converge stably and rapidly, which allows the aggregated latent features to be simultaneously reconstructed and adaptively classified. Compared with the state-of-the-art methods on the NEU-CLS, Data-Crack, and DAGM2007 datasets, our method achieves the most stable performance and the highest accuracy in different defect detection tasks.

Highly-squinted synthetic aperture radar (SAR) echo has the characteristic of severe range-azimuth coupling, requiring specialized imaging algorithms. Applications of compressed sensing in SAR imaging can effectively improve the resolution and other indicators. However, inaccurate manual parameters can affect the algorithm output. This article proposes an improved alternating direction method of multipliers (ADMM) for solving sparse reconstruction models under highly-squinted conditions. By adaptively adjusting the penalty parameter in ADMM via hyper-gradient descent (HD), the problem caused by inaccurate manual parameter is solved. Compared with matched filtering methods and other optimization methods, this method can suppress noise and speed up convergence. The effectiveness of the proposed method can be validated through the approximate observation of both simulated scenes and real scenes captured by the GF-3 SAR satellite.

In this paper, we propose a fence recognition method based on the ENet (Efficient neural Network) segmentation network to address the problems of traditional segmentation networks, which have poor performance in recognizing fences with a large range of scale variations and hollow structures.Firstly, a multi-scale attention fusion ENet segmentation network is designed, which is trained using the fence with obvious color features.Then, a morphological algorithm is used to process the predicted image to restore the fence segmentation results.The designed multi-scale attention fusion segmentation network performs better on fence datasets than traditional methods.In addition, the activation function Leaky_Relu6 further enhances the stability and generalization ability of the network.The experiments are conducted on 540 fence images from different construction sites, and the computed IoU is 90%.The processing speed is about 28 frames per second.The experimental results show that our proposed network outperforms traditional segmentation algorithms in fence recognition performance, and achieves robustness in different construction scenarios while meeting the requirements of both accuracy and speed.

Ship target recognition is a vital task in synthetic aperture radar (SAR) imaging applications. Although convolutional neural networks have been successfully employed for SAR image target recognition, surpassing traditional algorithms, most existing research concentrates on the amplitude domain and neglects the essential phase information. Furthermore, several complex-valued neural networks utilize average pooling to achieve full complex values, resulting in suboptimal performance. To address these concerns, this paper introduces a Complex-valued Convolutional Neural Network (CVGG-Net) specifically designed for SAR image ship recognition. CVGG-Net effectively leverages both the amplitude and phase information in complex-valued SAR data. Additionally, this study examines the impact of various widely-used complex activation functions on network performance and presents a novel complex max-pooling method, called Complex Area Max-Pooling. Experimental results from two measured SAR datasets demonstrate that the proposed algorithm outperforms conventional real-valued convolutional neural networks. The proposed framework is validated on several SAR datasets.

Welcome to the concluding issue of IEEE Transactions on Computational Social Systems (TCSS) for the year 2023. We would like to seize this opportunity to extend our heartfelt appreciation and congratulations to all for your exceptional dedication and unwavering support. We eagerly anticipate further collaboration to enhance the publication quality and expedite the review process of TCSS in the upcoming year 2024.

Pansharpening aims to generate a high-resolution multi-spectral (HR-MS) image given a paired panchromatic (PAN) image and low-resolution multi-spectral (LR-MS) image. Though existing pansharpening methods have made remarkable progress, the fusion pipeline does not fully adapt to the distinct characteristics of the PAN and LR-MS images. In this paper, to fully exploit the complementary modality of the two images, we propose a novel and efficient asymmetric bidirectional fusion network (ABFNet). The ABFNet consists of the two customized fusion modules with asymmetric architectures, which aim to reinforce the PAN and LR-MS images respectively. Specifically, the spectral colorization module recalibrates the scale and bias of the PAN features using weights generated by the LR-MS features, which aims to inject spectral information into the PAN features without breaking their spatial continuity. To transfer spatial details from the PAN features into the LR-MS features, the spatial restoration codebook module refines the LR-MS features with point-to-point restoration codebooks learned from the PAN features. By incorporating the two modules in multiple stages, ABFNet enjoys a high capability for capturing both spectral and spatial dependencies. Extensive experiments over multiple satellite datasets demonstrate the effectiveness of the proposed methods.

The key to solving the few-shot image classification problem is learning image category information from a handful of image samples. Few-shot image classification can easily cause overfitting problems that impact classification performance due to insufficient labeled data. In this paper, an attention mechanism-based relation network model for few-shot image classification is proposed. Inspired by classic methods of relation networks for few-shot learning, the attention mechanism in a convolutional block attention mechanism (CBAM) for metric learning is utilized in the feature extraction network. Then, an update strategy selecting a validation set during the training is adopted to reduce the possibility of overfitting. Through comparison experiments on different datasets, the results demonstrate that our model has better accuracy than traditional methods.

There has been a rising trend of solving few-shot image classification problems utilizing graph neural networks (GNN) in recent years. However, most GNN-based approaches fail to fully exploit relationships among samples, resulting in poor image classification. To address this problem, we propose a feature-enhanced GNN model appropriate for few-shot image classification tasks. The suggested method first exploits an efficient convolutional block to generate accurate feature maps, enhancing the expressivity of the feature extraction module. Then, our technique employs flexible combinatorial distance metric functions to compute the exact relation score between samples and determine image relevance by minimizing the matching cost. Moreover, a multilayer perceptron based on the residual structure with an attention mechanism is developed to produce focused feature representations, allowing the model to obtain selective and relevant information among the samples. The proposed model is evaluated on a supervised few-shot image classification task utilizing four benchmark datasets, with the corresponding results demonstrating that our model achieves a higher accuracy performance than traditional few-shot image classification methods.

Due to their powerful modeling and reasoning capabilities, graph neural networks have not only achieved adequate performance in unstructured data but, in recent years, their research interests in Euclidean data such as images have also been on the rise. In this context, the most common task is image classification, whose method, based on graph neural networks, is roughly divided into two stages. The first is the graph construction stage, where the images are converted into graph structure data (composed of a node-edge-node form). The second is the graph classification stage, where the processed graph data is loaded into the graph classification network for graph classification. Subsequently, the images classification results are obtained. However, nearly two problems are faced in this setting. One is that the graph construction stage takes a long time due to the considerable computation that is required in order to convert images into graph structured data, the other problem is related to the graph classification used in the graph classification stage. The number of layers in the network tends to be limited, usually 4 layers or less. Hence, the graph’s classification accuracy is affected to some extent. Subsequently, this study proposes a deep graph neural image classification model based on gSLIC, combining the attention mechanism to conduct related experiments, in order to prove that while the speed of graph construction is greatly improved, the image classification accuracy exceeds that of most existing models based on graph neural networks for image classification.

Tomographic synthetic aperture radar (TomoSAR) is an advanced SAR interferometric technique to retrieve 3-D spatial information. However, decorrelation effects degrade the quality of interferometric phases, resulting in errors in the reconstruction. In this paper, we propose a denoising method based on the unsupervised convolution neural network (CNN) with a loss function combining the deterministic descriptive regularization and total variation (TV) term. It can improve both the accuracy and completeness of the reconstructed 3-D point clouds, which is verified by experiments on simulated and real SAR images.

Urban region function recognition plays a vital character in monitoring and managing the limited urban areas. Since urban functions are complex and full of social-economic properties, simply using remote sensing (RS) images equipped with physical and optical information cannot completely solve the classification task. On the other hand, with the development of mobile communication and the internet, the acquisition of geospatial big data (GBD) becomes possible. In this paper, we propose a Multi-dimension Feature Learning Model (MDFL) using high-dimensional GBD data in conjunction with RS images for urban region function recognition. When extracting multi-dimension features, our model considers the user-related information modeled by their activity, as well as the region-based information abstracted from the region graph. Furthermore, we propose a decision fusion network that integrates the decisions from several neural networks and machine learning classifiers, and the final decision is made considering both the visual cue from the RS images and the social information from the GBD data. Through quantitative evaluation, we demonstrate that our model achieves overall accuracy at 92.75%, outperforming the state-of-the-art by 10% percent.

Synthetic aperture radar (SAR) images have been widely used in water body information extraction. However, SAR images suffer from speckles and the additive noise, which affect the performance of automatic information extraction. Thus, we propose the nonconvex-nonlocal total variation (NLTV) regularization to suppress speckles and the additive noise, and improve the performance of water body information extraction using the enhanced images. Experiments using Qilu-1 (QL-1) SAR data verify the effectiveness of the method.

High resolution and high quality are now the requirements in synthetic aperture radar (SAR) research. The sliding spotlight mode can obtain high azimuth resolution because of its large azimuth bandwidth. Group sparse penalty can effectively suppress azimuth ambiguities to improve image quality. Generalized mini-max concave (GMC) penalty is a kind of nonconvex penalty, which is widely used in SAR imaging. In this paper, a novel sliding spotlight SAR imaging method based on group sparsity and nonconvex regularization is proposed. Compared with matched filtering method, the proposed method can suppress noise and azimuth ambiguities. Both simulations and Qilu-1(QL-1) real SAR data experiments verify the effectiveness of the proposed method.

Anomaly detection technology relying on memory reconstruction leverages the difference in reconstruction errors between the normal and abnormal frames to achieve superior detection performance. However, there are still some challenges with this technology. First, the memory has insufficient representation capacity for features. Second, there is a contradiction between feature fusion and reconstruction. As feature fusion copies the abnormal patterns into the reconstructed frames, the abnormal frames are effectively reconstructed, reducing the detection performance. In response to these challenges, we use a memory update threshold to improve the representational power of memory. We also propose a dual-encoder anomaly detection model to restrict anomaly feature propagation. Experiment results demonstrate the effectiveness and robustness of our approach.

Abstract Small objects only occupy a few pixels in an image, which results in low performance of small object detection for existing object detection algorithms. Therefore, the authors propose a dense sampling and detail enhancement network (DSDE‐Net) to address this issue. The network contains a dense sampling module used to increase the resolution of feature maps and expand the receptive field, which includes an atrous spatial pyramid pooling network and a coordinate attention mechanism to systematically process feature maps. Simultaneously, the authors introduce a detail enhancement branch that contains edge and detailed information to generate detailed enhancement feature maps through Gaussian filtering to compensate for the loss of small object information that occurs in the feature extraction process. The experimental results demonstrate that the proposed network outperformed related methods. Compared with the state‐of‐the‐art algorithm DetectoRS, it effectively achieves approximately 4.6% improvement on the minicoco2021 dataset and 4.2% improvement on the remotely sensed dataset VisDrone.

Frequency-modulated continuous wave synthetic aperture ladar (FMCW-SAL) is an important remote sensing observation method. Sliding spotlight FMCW-SAL can obtain high resolution and wide observation coverage in azimuth simultaneously. This work models and simulates sliding spotlight FMCW-SAL signal. FMCW-SAL beam scanning leads to spectrum aliasing in the azimuth direction, and the linear and second-order position offset errors are introduced in the range direction. Compared with sliding spotlight FMCW synthetic aperture radar (SAR), the linear position offset error is more serious and must be corrected. However, there is no linear position offset correction in previous research studies in sliding spotlight FMCW-SAR. An improved sliding spotlight FMCW-SAL imaging algorithm is proposed for the above problems here. This algorithm includes two new phase compensation factors, which can remove Doppler ambiguity and correct first-order, second-order position offset errors in range simultaneously. Simulation experiments verify the effectiveness of the signal model and the imaging algorithm.

Building segmentation of synthetic aperture radar (SAR) images is a challenging task that has not been solved well. High-resolution interferometric SAR (InSAR) images can provide delicate textures and interferometric phase images useful for building segmentation. However, current semantic segmentation networks in computer vision cannot be directly applied in InSAR building segmentation tasks to get good results because of the InSAR images’ particularity. In this article, we present a novel complex-valued convolutional and multifeature fusion network (CVCMFF Net) specifically for building semantic segmentation of InSAR images. This CVCMFF Net not only learns from the complex-valued SAR images but also considers multiscale and multichannel feature fusion. It can effectively segment the layover, shadow, and background on both the simulated InSAR building images and the real airborne InSAR images. The segmentation performance of CVCMFF Net is significantly improved compared with those of other state-of-the-art networks. By feature visualization, the feature extraction rule and feature fusion mechanism of the network are explored. We hope that the proposed network can be beneficial to InSAR phase filtering, phase unwrapping, and information extraction in urban areas.

Land use and land cover (LULC) classification plays a significant role in Earth observation tasks. Nowadays, we can observe the same scene with multiple heterogeneous sensors. Combining diverse information therein for multisource joint classification has become a promising research topic in the remote sensing community. For example, the fusion of hyperspectral image (HSI) and lidar detection and ranging (LiDAR) data has been under active research. The current methodology for HSI and LiDAR joint classification tends to ignore the topological relationship between pixels, limiting the effectiveness of feature extraction and fusion. Another obstacle to satisfactory performance is the scarcity of annotated data. To overcome the above challenges, this article proposes a multisource attention network called MAGE to improve the collective classification. We use a semi-supervised graph transductive module to underline the relevance among pixels by explicitly constructing a multimodal adjacency matrix. Specifically, MAGE designs a self-supervised feature extraction module for pre-training, mitigating the dependence on annotated samples and alleviating the common overfitting and over-smoothing problems encountered by the deep graph neural network (GNN). The experimental results of three standard datasets, i.e., MUUFL, Trento, and Houston, demonstrate the effectiveness of the proposed approach. In particular, MAGE achieves an overall accuracy of 95.26% and an average accuracy of 96.27% on the challenging MUUFL dataset, surpassing the state-of-the-art methods. The code and models are publicly available at https://github.com/d1x1u/MAGE.

There are shadows existing in traditional SAR tomography (TomoSAR) 3D imaging results, which bring difficulties in application of TomoSAR. However, TomoSAR with multi-azimuth observations can be applied to address this problem. In this paper, a panoramic 3D reconstruction method for TomoSAR based on multi-azimuth observations will be introduced. Firstly, 2D images are achieved with backprojection (BP) algorithm on the ground plane. Secondly, 3D reconstruction of TomoSAR is realized with orthogonal matching pursuit (OMP). After coordinate transformation, the panoramic 3D reconstruction results of TomoSAR based on multi-azimuth observations are achieved with registration of point clouds, in which iterative closest point algorithm (ICP) is effectively applied. Panoramic 3D reconstruction results of airborne TomoSAR experimental data can validate correctness and effectiveness of our method.

Supraglacial lakes have a great impact on the mass balance and dynamics of Greenland ice sheet. While the current studies on these lakes mainly utilize optical observation data, the validity is poor, and it is impossible to conduct spatio-temporal analyses. This study provides an automatic extraction model for supraglacial lake, using Synthetic Aperture Radar (SAR) data. By processing Sentinel-1 SAR dual-polarized imagery, a train dataset of 2664 image patches are formed. We use U-Net for segmentation and the associated Dice coefficient could reach higher than 95%. Besides, the terrain shadow is removed by DEM. The results are validated by supraglacial lake detection using Landsat 8. We will integrate more training data and optimize the feasibility of the model in future work.

The azimuth multichannel SAR is competent to achieve high-resolution and wide-swath (HRWS) imaging. For some spaceborne multichannel SAR systems, the pulse repetition frequency (PRF) of each channel at some beam positions is less than that of uniform sampling, hence leading to the azimuth ambiguities in the recovered images. In this paper, a novel azimuth ambiguity suppression method for multichannel SAR imaging based on$L_{2,q}$regularization$(0 < q\leq 1)$is proposed. First, by analyzing the reasons of azimuth ambiguities in multichannel SAR, we establish the imaging model different from that in single-channel SAR system. Second, we extend the$L_{2,q}$regularization from single-channel SAR system to multichannel SAR system and develop the proposed method. Finally, we demonstrate the effectiveness of the proposed method for non-sparse scenarios. Simulations and Gaofen-3 real data experiments are carried out to verify the validity of proposed method.

Abstract Speckle reduction is still an intractable task in ultrasound imaging field. Ultrasound speckle is usually described as multiplicative noise with its statistics following a Rayleigh or Gaussian distribution. To employ these two distributions effectively, the authors attempt to describe ultrasound speckle using a Weibull distribution, because it can include the Rayleigh distribution as a special case and also approximate a Gaussian distribution by varying its shape and scale parameters. The authors’ contribution in this paper is to propose a Weibull‐distribution‐based hybrid total variation (WHTV) method to reduce ultrasound speckle. The WHTV energy functional is convex and consists of a new data fidelity term and a new regularization term. The former is derived from the multiplicative Weibull model of ultrasound speckle based on the maximum likelihood criterion. The latter is a new edge‐weighted combination of the first‐ and second‐order total variation, with the advantage of preserving edges while alleviating the staircase effects. The minimization of the WHTV energy functional is implemented by the split Bregman algorithm. Experimental results on synthetic and real ultrasound images have demonstrated not only that the Weibull distribution is a better fitting model for the statistics of ultrasound speckle than other distributions such as Rayleigh, Gaussian, Gamma, and Nakagami, but also that the proposed WHTV method can achieve better despeckling performance than several state‐of‐the‐art variational methods.

Spotlight synthetic aperture radar (SAR) achieves very high-resolution (VHR) images by steering the azimuth beam during the formation of the synthetic aperture. In practice, the steering is implemented through discrete azimuth beam switching. Then, phase shifts can occur between the adjacent beams due to the error of the antenna pattern. In addition, the troposphere introduces a beam-angle-dependent delay to the echo. Those undesired phase shifts and delays cause phase errors in the received echo and result in image quality deterioration. In this letter, an algorithm, called the Newton entropy minimization (N-EM), is proposed to estimate and compensate the phase errors caused by the discrete azimuth beam steering for the spotlight SAR data. Combining with the subaperture imaging approach of the spotlight SAR, the algorithm estimates the phase offsets between each couple of the adjacent beams based on the minimum entropy criterion. The analytic expression, which is a nonlinear equation, is developed for the optimal estimation. Then, the Newton's method is employed to solve the equation. The real spaceborne SAR (both the staring and sliding spotlight modes) data processing results demonstrate the efficiency and accuracy of the proposed algorithm.

Inverse Synthetic Aperture Ladar (ISAL) can continuously capture images of the target in complex motion, which provides an opportunity for the 3-D reconstruction. However, the existing 3-D imaging methods need to meet some specific conditions, including multiple isolated prominent scattering points in the image, known or parametric motion of the target, and 360° observation of the target. To relax these restrictions, a novel 3-D reconstruction method is proposed. In this method, the corner of the target is first detected and used as a reference point. Then, the positioning accuracy of these fiducial points is greatly improved by the relationship between the radial range and velocity. Finally, the 3-D reconstruction is realized by the singular value decomposition of the measurement matrix. The data obtained from the outdoor experiment are used to test the proposed method. By this process, a series of images of a satellite model which is 2 km away are acquired. The size and shape errors of the reconstruction are less than 10%. The result demonstrates the effectiveness of the 3-D reconstruction method.

Circular SAR (CSAR) is a typical 3-D imaging model of SAR. 3-D back-projection (BP) algorithm is a common time-domain 3-D imaging method for CSAR. However, 3-D imaging with back-projection algorithm has high algorithm complexity and low efficiency for processing pulse by pulse and grid by grid. This paper proposes a new fast 3-D imaging method for CSAR based on 3-D back-projection algorithm. In this method, 3-D interpolation and phase compensation operations can be transformed into 1-D interpolation and phase compensation operations and matrix searching operations with the construction of a geometric interpolation kernel. This proposed method can greatly improve imaging efficiency of CSAR when the error range allows. The simulated experimental results can prove the correctness and effectiveness of the proposed method.

The micro-Doppler modulation of the radar echo of the drone's rotor reflects the micro-movement characteristics of the target. Accurate estimation of the length and rotation frequency of an unmanned aerial vehicle (UAV) rotor is of great significance for target identification and classification in radar echoes. Firstly, this paper proposes a method of optimal estimation based on concentration of time-frequency rotation domain (CTFRD), in the time-frequency rotation domain of a multi-component micro-Doppler signal, under the FMCW radar system. Secondly, in the scene where the drone rotor rotates at a constant speed or at a uniform acceleration, the proposed method realizes the accurate estimation for multicomponent micro-motion feature parameters. Compared to traditional methods, it is also very robust in low signal-to-noise ratio (SNR) environments. Finally, the effectiveness of the proposed method is verified by simulations and real-world scenarios. Index Terms- Micro-Doppler, Concentration of time-frequency rotation domain, Parameter estimation, Target identification.

It has become a field of intensive research to exploit SAR tomography to reconstruct a 3-D model of the buildings. However, in multiple-input multiple-output (MIMO) SAR tomography, the double-bounce reflections of the building corner will cause symmetric virtual scatterers, affecting both the scattering coefficient and structure of the 3-D model. To solve this problem, the building corner reflection is discussed and compressive sensing (CS)-based corner reflection suppression (CSCRS) is proposed. In the final part of this letter, the effectiveness of the proposed method is validated using array InSAR data. It is found that the proposed method leads to considerable improvements with regard to suppression ratio and reconstruction accuracy.

In recent years, the sparse signal processing technique has shown significant potential in synthetic aperture radar (SAR) imaging, such as image performance improvement and downsampled data-based image recovery. However, due to the huge computational complexity needed, the existing sparse SAR imaging methods, such as conventional observation matrix-based and azimuth-range decouple-based algorithms, are not able to achieve real-time processing, especially for the large-scale scenes, which seriously restricts its application in some fields, e.g., real-time monitoring and early warning. To solve this problem, this article presents a novel real-time sparse SAR imaging method, which can get a similar image performance to that obtained by the existing sparse imaging methods, to reduce the computational complexity to the same order as that required by matched filtering (MF)-based algorithms. This means that with the proposed method, real-time data processing for practical large-scale scene sparse reconstruction becomes possible. Experimental results based on simulated and real data along with a performance analysis are presented to validate the proposed real-time sparse imaging method.

Sparse signal processing theory has been applied to SAR imaging. The estimation of sparsity is crucial for sparse SAR imaging. But the true value of sparsity is unknown. Adaptive parameter estimation for sparse SAR imaging can achieved by the automatic regularization parameter estimating methods. However, these methods are deduced based on measurement matrix, which will cause huge computational and memory costs. Also, the adaptive estimated sparsity is often greater than the true value due the noise and sidelobes. In this paper, we propose improved adaptive parameter estimation method for sparse SAR imaging. The complex-image-based sparse SAR imaging is adopted to pre-estimate the parameter. Then, azimuth-range decouple operators are introduced into parameter estimation method. Simulation and real data experimental results show the effectiveness of the proposed method.

In this paper, basis pursuit denoising (BPDN) is applied to extract three-dimensional scattering centers of complex radar targets. Since the distributions of scattering centers are usually sparse in the high-frequency optics region, the valid backscattering coefficients of radar targets can be obtained with the undersampled measurements in azimuth and elevation. It significantly reduces the time for measuring echo signals, which promotes the efficiency of radar systems. However, due to the great computation load and memory cost caused by the matrix-vector products, most of the sparse reconstruction algorithms are not suitable for the process of scattering center extraction. To solve this problem, we derive accelerated operators on the basis of support set and filtered backprojection. By combining the sparse solver SPGL1 with the proposed operators, the new algorithm not only reduces the consumption of system resources, but also estimates the backscattering coefficients accurately. Additionally, the experimental results and analysis demonstrate that the proposed technique possesses high data compression ratio and small radar cross section (RCS) reconstruction error.

This study proposes a new, simple but effective technique to detect and restore colour cast images, named modified grey world method. This method detects colour cast images of outdoor surveillance videos by computing the values in the YUV colour space, which makes it much easier than classic methods. Specific colour cast can be found out by calculating the hue values. Additionally, this method can detect not only simple colour cast images but also multiple colour cast images simultaneously. To detect and restore a colour cast image, the authors first remove all grey pixels and separate it into multiple parts with a maze‐solving algorithm. Then, they compute the YUV colour values of each part. If the values are too high or too low, this part of the input image is designated as a colour cast. Finally, they carry out a restoration procedure, in which they calculate weights by matching average colour value with a grey reference value in YUV colour space. This method has been tested in the Safety City surveillance system in Wuhan city, China. The results show that the proposed method leads to better results in detecting and restoring colour cast imaging than classic methods in outdoor surveillance videos.

The inverse synthetic aperture ladar (ISAL) is an important method for observation and imaging of space targets. Here, a 3-D ISAL imaging algorithm is proposed for spinning targets such as space debris. Since laser wavelength is 4-5 orders of magnitude smaller than that of microwave, the Doppler frequency caused by target motion is more pronounced in ISAL. Doppler frequency modulation rates can be estimated by the fractional Fourier transform with respect to azimuth slow time even when the rotation angle is small such that scattering centers do not migrate through a range cell. Then, slant range, Doppler frequency, and Doppler frequency modulation rates form a 3-D space. The angular velocity and the incident angle can be estimated by the position relationship between the scattering centers in two observations. After image scaling, the 3-D shape and size of the target can be obtained. The 3-D structure of the target in the simulation experiment is accurately reconstructed. Monte Carlo experiments are conducted to discuss the effect of the signal-to-noise ratio and observation time on the algorithm. Finally, the effectiveness and robustness of the algorithm are verified.

Sparse signal processing has been applied in synthetic-aperture radar (SAR) imaging. As a typical sparse reconstruction model, L1regularization often underestimates the intensities of the targets. The estimated radar cross section (RCS) is related to the pixel intensity. Thus, the linear relationship between the targets' intensities cannot kept. The underestimation will also cause radiometric errors and affect the quantitative use of the SAR data. In this letter, we present a SAR imaging method based on generalized minimax concave (GMC) penalty. GMC is a nonconvex penalty and its cost function is convex. GMC can avoid the underestimation of pixel intensity. In the iteration, the azimuth-range decouple operators are used to avoid the huge memory and computational costs. The performance of the proposed method is verified using real data.

Single image super-resolution is an effective way to enhance the spatial resolution of remote sensing image, which is crucial for many applications such as target detection and image classification. However, existing methods based on the neural network usually have small receptive fields and ignore the image detail. We propose a novel method named deep memory connected network (DMCN) based on a convolutional neural network to reconstruct high-quality super-resolution images. We build local and global memory connections to combine image detail with environmental information. To further reduce parameters and ease time-consuming, we propose downsampling units, shrinking the spatial size of feature maps. We test DMCN on three remote sensing datasets with different spatial resolution. Experimental results indicate that our method yields promising improvements in both accuracy and visual performance over the current state-of-the-art.

Discriminative tracking methods can achieve state‐of‐the‐art performance by considering tracking as a classification problem tackled with both object and background information. As a high efficient discriminative tracker, compressive tracking (CT) has attracted much attention recently. However, it may easily fail when the object suffers from long‐term occlusions, and severe appearance and illumination changes. To address these issues, the authors develop a robust tracking framework based on CT by considering balanced feature representation as well as dual‐mode classifier construction. First, the original measurement matrix of CT works as a dominated texture feature extractor. To obtain a balanced feature representation, they propose to induce a complementary measurement matrix by considering both texture and colour features. Then, they develop two classifiers (dual mode) by using previous and current sample sets, respectively, and subsequently combine them into one ensemble classifier to track the target, which can help to avoid tracking failure suffering from severe appearance changes and long term occlusion. Moreover, they propose a classifier updating schema to prevent the inclusion of unsatisfied positive samples by predicting the occlusions with their ensemble classifier. The extensive experiments demonstrate the superior performance of their tracking framework under various situations.

Synthetic aperture radar (SAR) is a widely used active high-resolution microwave imaging technique that has alltime and all-weather reconnaissance ability. Compared with traditionally matched filtering (MF)-based methods, Lq(0 ≤ q ≤ 1) regularization technique can efficiently improve SAR imaging performance e.g., suppressing sidelobes and clutter. However, conventional Lq-regularization-based SAR imaging approach requires transferring the 2-D echo data into a vector and reconstructing the scene via 2-D matrix operations. This leads to significantly more computational complexity compared with MF, and makes it very difficult to apply in high-resolution and wide-swath imaging. Typical Lqregularization recovery algorithms, e.g., iterative thresholding algorithm, can improve imaging performance of bright targets, but not preserve the image background distribution well. Thus, image background statistical-property-based applications, such as constant false alarm rate (CFAR) detection, cannot be applied to regularization recovered SAR images. On the other hand, complex approximated message passing (CAMP), an iterative recovery algorithm for L1regularization reconstruction, can achieve not only the sparse estimation of the original signal as typical regularization recovery algorithms but also a nonsparse solution simultaneously. In this paper, two novel CAMP-based SAR imaging algorithms are proposed for raw data and complex radar image data, respectively, along with CFAR detection via the CAMP recovered nonsparse result. The proposed method for raw data can not only improve SAR image performance as conventional L1regularization technique but also reduce the computational cost efficiently. While only when we have MF recovered SAR complex image rather than raw data, the proposed method for complex image data can achieve a similar reconstructed image quality as the regularization-based SAR imaging approach using the full raw data. The most important contribution of this paper is that the proposed CAMP-based methods make CFAR detection based on the regularization reconstruction SAR image possible using their nonsparse scene estimations, which has a similar background statistical distribution as the MF recovered images. The experimental results validated the effectiveness of the proposed methods and the feasibility of the recovered nonsparse images being used for CFAR detection.

Haze significantly degrades the visibility for ship navigation and traffic monitoring in China's inland waterways. In this study, the authors propose a novel haze‐removal method based on sky segmentation and dark channel prior to restore images. Sky segmentation is accomplished by using robust image matting and region growth algorithms. Then, the average image intensity of the sky region is chosen as the atmospheric light value to address the defect of dark channel prior. Experimental results show that their method can restore inland waterway images effectively; restored images are more natural and smoother than those obtained by the state‐of‐the‐art haze removal algorithms.

Predicting breast cancer risk has long been a goal of medical research in the pursuit of precision medicine. The goal of this study is to develop novel penalized methods to improve breast cancer risk prediction by leveraging structure information in electronic health records. We conducted a retrospective case- control study, garnering 49 mammography descriptors and 77 high- frequency/low-penetrance single-nucleotide polymorphisms (SNPs) from an existing personalized medicine data repository. Structured mammography reports and breast imaging features have long been part of a standard electronic health record (EHR), and genetic markers likely will be in the near future. Lasso and its variants are widely used approaches to integrated learning and feature selection, and our methodological contribution is to incorporate the dependence structure among the features into these approaches. More specifically, we propose a new methodology by combining group penalty and $\ell^p$ ($1\leq p\leq2$) fusion penalty to improve breast cancer risk prediction, taking into account structure information in mammography descriptors and SNPs. We demonstrate that our method provides benefits that are both statistically significant and potentially significant to people's lives.

Due to the great computation load and memory cost of the matrix-vector multiplication, the sparse reconstruction algorithms are severely limited in the applications of radar imaging with real data. In order to solve this problem, we construct a backprojection-based range-azimuth decoupled operator (BP-RADOp) and combine the complex approximate message passing algorithm (CAMP) with it. We call this algorithm BP-CAMP in this letter. Since BP-RADOp retains the merits of the backprojection method entirely (i.e., perfect motion compensation for any flight path, precise focus for arbitrarily wide bandwidths and integration angles, low artifact levels, unlimited scene size, and strictly local processing), it has universal applicability in comparison with the other decoupled operators deduced from the fast Fourier transform-based image formation algorithms. The theoretical analysis indicates when BP-CAMP and CAMP are both used to reconstruct large-scale observed scenes; the former has lower computation load and memory cost than the latter. Meanwhile, it is demonstrated that BP-CAMP achieves high-quality synthetic aperture radar imaging with undersampled echo data, and it is as robust as CAMP to additive noise by the simulations and real data processing.

The displaced phase center antenna (DPCA) technique is a classical method for achieving high-resolution wide-swath synthetic aperture radar (SAR) imaging. For optimum performance, the pulse repetition frequency (PRF) of DPCA SAR systems should satisfy the azimuth uniform sampling condition as far as possible. However, this rigid PRF selection may conflict with the timing diagram for some incidence angles, which usually results in a nonuniform sampling of the synthetic aperture. According to the sparse signal processing theory, this letter proposes a novel DPCA imaging algorithm for the nonuniform displaced phase center sampling. By combining the DPCA data processing operator based on a filter bank with the ℓqregularization scheme, the algorithm can efficiently recover the backscattering coefficients of the observed scene. The experimental results have shown that it is capable of resolving ambiguity and suppressing clutter effectively and is meanwhile insensitive to additive noise.

Recently, the sparse reconstruction algorithms (SRAs) based on compressive sensing (CS) have been applied in the fields of synthetic aperture radar (SAR) imaging and show plenty of potential advantages. However, due to the great computational complexity and memory cost caused by matrix-vector multiplications, most of these algorithms are not suitable to reconstruct large-scale observed scenes. To solve this problem, we construct a backprojection based imaging operator, and introduce it to the complex approximate message passing algorithm (CAMP). The new image formation algorithm is called BP-CAMP in this paper. Compared with the approximated observation methods deduced from the FFT-based imaging technology, BP-CAMP is not limited by observation models of the radar and motion modes of the platform, and it therefore possesses universal applicability. By the simulations and real data processing, the experimental results show that BP-CAMP has lower computational complexity and memory cost than CAMP, and also achieves SAR imaging with under-sampled echo data.

Microwave imaging, or synthetic aperture radar (SAR) shows its remarkable importances in various fields of remote sensing. Modern SAR system usually comes with high imaging resolution and wide mapping swath. This brings difficulties to the future development of SAR system. As a solution, the concept of SAR imaging under sparse constraint, or sparse microwave imaging radar is suggested, which is mainly the idea of introducing the sparse signal processing theory to the radar imaging. Under the sparse constraint, this technique could bring us benefits including better imaging performance e.g. lower ambiguity, higher resolution, lower side lobe and lower system complexity [1, 2, 3, 4].

Compressed sensing (CS) techniques can reduce the sampling rates required in synthetic aperture radar (SAR). However, it is difficult to use the restricted isometry property to theoretically analyze the performance. Therefore, in this letter, information theory is applied to set necessary bounds on sampling ratios in CS-based SAR imaging. The system is viewed as a multi-input/multi-output (MIMO) channel, with information capacity quantified for a given measurement matrix and signal-to-noise ratio (SNR). According to the source-channel coding theorem, the lower bound of the sampling ratios is derived in terms of sparsity ratio, SNR, bandwidth, and radar pulse duration. Simulation studies are performed to test and analyze the information-theoretical bounds.

We introduce a total variation (TV) regularization model for synthetic aperture radar (SAR) image despeckling. A dual-formulation-based adaptive TV (ATV) regularization method is applied to solve the TV regularization. The parameter adaptation of the TV regularization is performed based on the noise level estimated via wavelets. The TV-regularization-based image restoration model has a good performance in preserving image sharpness and edges while removing noises, and it is therefore effective for edge preserve SAR image despeckling. Experiments have been carried out using optical images contaminated with artificial speckles first and then SAR images. A despeckling evaluation index (DEI) is designed to assess the effectiveness of edge preserve despeckling on SAR images, which is based on the ratio of the standard deviations of two neighborhood areas of different sizes of a pixel. Experimental results show that the proposed ATV method can effectively suppress SAR image speckles without compromising the edge sharpness of image features according to both subjective visual assessment of image quality and objective evaluation using DEI.

Bayesian network structures are usually built using only the data and starting from an empty network or from a naïve Bayes structure. Very often, in some domains, like medicine, a prior structure knowledge is already known. This structure can be automatically or manually refined in search for better performance models. In this work, we take Bayesian networks built by specialists and show that minor perturbations to this original network can yield better classifiers with a very small computational cost, while maintaining most of the intended meaning of the original model.

Airborne downward-looking sparse linear array 3-D synthetic aperture radar operates nadir observation and obtains the 3-D microwave scatter information of the observed scene. A polar format algorithm (PFA) with space-variant wave-front curvature phase error compensation is presented. A 3-D image in polar coordinate can be obtained with the proposed PFA, and a 3-D image in Cartesian coordinate can be obtained with interpolation. The proposed PFA possesses the advantages of high precision, low memory requirement, and low computational complexity. The focus performance of the proposed PFA is validated by 3-D distributed scene simulation with an airborne X-band digital elevation model and a P-band circular SAR image of the same area as simulation scene input.

When mammography reveals a suspicious finding, a core needle biopsy is usually recommended. In 5% to 15% of these cases, the biopsy diagnosis is non-definitive and a more invasive surgical excisional biopsy is recommended to confirm a diagnosis. The majority of these cases will ultimately be proven benign. The use of excisional biopsy for diagnosis negatively impacts patient quality of life and increases costs to the healthcare system. In this work, we employ a multi-relational machine learning approach to predict when a patient with a non-definitive core needle biopsy diagnosis need not undergo an excisional biopsy procedure because the risk of malignancy is low.

In this paper, we introduce a total variation (TV) regularization model for SAR image despeckling. A dual formulation based adaptive total variation (ATV) regularization method is applied to solve the TV regularization. The parameter adaptation of the TV regularization is performed based on the noise level estimated via wavelets. The TV regularization based image restoration model has a good performance in preserving image sharpness and edges while removing noises and it is therefore effective for edge preserve SAR image despeckling. Experiments have been carried out using optical images contaminated with artificial speckles first and then SAR images. An evaluation index is designed to assess the effectiveness of edge preserve despeckling on SAR images, which is based on the ratio of the standard deviations of two neighborhood areas of a pixel with different sizes. Experimental results show that the proposed method can effectively suppress SAR image speckles without compromise the edge sharpness of image features according to both subjective visual examination and objective evaluation indices of image quality.

Synthetic aperture radar (SAR) images are inherently affected by multiplicative speckle noise, which will degrade the human interpretation and computer-aided scene analysis. In this paper, we propose a novel Bayesian multiscale method for SAR image despeckling in the non-homomorphic framework. To address the multiplicative nature, we first make the speckle contribution additive by a linear decomposition. Then, in the stationary wavelet transform domain, a two-sided generalized Gamma distribution (GTD) is introduced as a prior to capture the heavy-tailed nature of wavelet coefficients of the noise-free reflectivity. By exploiting this prior together with a Gaussian likelihood, an analytical wavelet shrinkage function is derived based on maximum a posteriori criteria, which further adopts heterogeneity-adaptive thresholding technique to achieve better estimates of noise-free wavelet coefficients. Moreover, a pilot-signal-assisted strategy is proposed to estimate the parameters of two-sided GTD with the estimator based on second-kind cumulants. Finally, experimental results, carried out on the synthetic and actual SAR images, are given to demonstrate the validity of the proposed despeckling method.

Range cell migration (RCM) correction and azimuth spectrum being contained entirely in baseband are critical for ground moving-target imaging (GMTIm). Without the azimuth spectrum entirely contained within baseband and a proper RCM correction, the image will be defocused, or artifacts may appear in the image. An instantaneous-range-Doppler algorithm of GMTIm based on deramp-keystone processing is proposed. The main idea is to focus all the targets in the scene at an arbitrarily chosen azimuth time. With our proposed algorithm, RCMs of all targets in the scene are removed without a priori knowledge of their accurate motion parameters. The targets with azimuth spectrum not entirely in baseband, i.e., azimuth spectrum within an ambiguous pulse repeating frequency (PRF) band or spanning neighboring PRF bands, can also be effectively dealt with simultaneously. Theoretical analysis shows that no interpolation is needed. The simulated and real data are used to validate the effectiveness of this method.

For different applications with different requirements, many synthetic aperture radar (SAR) modes have been developed in the literature, such as, Terrain Observation by Progressive Scans (TOPS) SAR and sliding spotlight SAR. In this paper, we call TOPS SAR, sliding spotlight SAR, and spotlight SAR as beam steering SAR (BS-SAR for short). Comparing with stripmap SAR, BS-SAR can obtain a wide diversity of resolutions by increasing or reducing the azimuth synthetic time. Traditional polar formation algorithm (PFA) is an efficient algorithm which is mainly developed for spotlight SAR. The PFA has been validated to obtain well-focused results of raw data. In this paper, we extend the traditional PFA to process sliding spotlight SAR and TOPS SAR data, and we call it generalized PFA (GPFA). Comparing with the traditional PFA, GPFA contains a different azimuth deramping function and an additional azimuth scaling operation. The simulated and real data are used to validate the effectiveness of this method.

Terrain Observation by Progressive Scans (TOPS) synthetic aperture radar (SAR) and spotlight SAR are advanced SAR imaging modes for wide range swath and high resolution. In order to obtain a wider range coverage, azimuth multichannel is introduced in the literature. Since the azimuth bandwidth of beam steering SAR (BS-SAR; spotlight SAR, sliding spotlight SAR, or TOPS SAR) is much greater than that of a stripmap SAR, a signal reconstruction algorithm used for multichannel stripmap SAR may not be effective for multichannel BS-SAR. In this paper, a multichannel full-aperture azimuth processing algorithm is proposed for a BS-SAR. The key of this algorithm lies in the beam and the azimuth bandwidth compressions of multichannel signals in the Doppler-array and slow time-angle planes, respectively. Through compression processing, the beamwidth and the azimuth bandwidth are smaller than the available angle and equivalent pulse repeating frequency , respectively. Then, an improved post-Doppler STAP method is proposed to recover a 2-D spectrum. With the recovered signal, further processing can be utilized to focus the multichannel signal. Simulation and real data results show the effectiveness of the proposed algorithm.

Many imaging algorithms for different modes, such as, stripmap synthetic aperture radar (SAR), spotlight SAR, sliding spotlight SAR, and terrain observation by progressive scans (TOPS) SAR, of SAR have been studied. This paper is to obtain a unified focusing algorithm (UFA) for these SAR modes based on fractional Fourier transform. By defining the rotation-center range, the stripmap SAR and spotlight SAR can be treated as special cases of sliding spotlight SAR or TOPS SAR. Then, a parameterized focusing algorithm determined by the rotation-center range is presented. Data of each mode can be focused by utilizing UFA and selecting parameters or rotation angles. Some application aspects of UFA are also analyzed. Simulation and real data results are presented to validate the analysis and the proposed method.

The “stop-go” approximation is widely used for the processing of synthetic aperture radar (SAR) data, and the error brought by this assumption can be negligible for most SAR systems. However, for the SAR on a high-speed platform, with the increasing requirements on high-resolution imaging, the error may be intolerable for SAR imaging. In this case, the radar motion within a pulse repetition interval should be taken into account for the echo model and imaging algorithm. In this paper, according to the geometric configuration of the SAR working process, an accurate echo model is presented. By comparing the “stop-go” echo (which denotes the echo based on the “stop-go” approximation in this paper) with the accurate echo, the error brought by the “stop-go” approximation is introduced, and the intolerable error is shown in a reference system. A spotlight imaging algorithm based on the accurate echo is given and is well supported by the simulation results.

The displaced phase center antenna (DPCA) synthetic aperture radar (SAR) has the potential to achieve high azimuth resolution and wide swath. Its pulse repletion frequency (PRF) has to be selected such that SAR platform moves just one half of its total antenna length between subsequent radar pulses. If this condition is not satisfied, there will be nonuniform sampling in azimuth and azimuth ambiguities will appear when traditional imaging algorithms based on matched filter are used. We propose an innovative imaging algorithm based on compressed sensing (CS) which can reconstruct the scene well even though this rigid condition is not satisfied.

In this paper, a novel synthetic aperture radar (SAR) imaging method based on L1/2regularization is proposed. Our method implements SAR imaging from compressed measurements with high resolution, enhanced features, reduced sidelobes and suppressed artifacts. Real SAR data experiments are implemented to demonstrate the outperformance of our method. The experiment results demonstrate that our method needs far below the traditional Nyquist rate to guarantee successful imaging. Compared to the prevalent L1regularization-based methods, there is a significant reduction of the sampling rate for SAR imaging. The sampling rate used by our method is about half of the L1regularization-based methods in the real SAR data experiments.

This letter presents a new algorithm for squint circular synthetic aperture radar (SAR) (CSAR) imaging, which is an extension of the well-known range migration algorithm. Due to the circular trajectory, the spatial frequency domain data of squint CSAR cannot be readily obtained via fast Fourier transform, as conventional SAR with straight path does. This method first employs along-track varying system kernels and filters to transform the raw data to the polar spatial frequency domain. Then, it uses an interpolation algorithm to convert the polar samples into rectilinear samples. Implementation aspects, including sampling criteria, resolutions, and computational complexity, are also assessed in this letter. The proposed algorithm is validated both numerically and experimentally.

The aperture of 360° gives circular synthetic aperture radar (SAR) (CSAR) the capability to detect hidden target when its orientation is unknown. Subwavelength resolution can also be achieved when the target in the spotted area is observed under a complete circular aperture. Furthermore, the aspect angle diversity inherent to the circular trajectory makes possible a 3-D target reconstruction. However, the latter two potentials require certain target reflectivity homogeneity. For a highly directive scatterer, it has no resolving ability in the direction normal to the data collection plane. In this letter, a new interferometric CSAR method is presented to enhance the tomographic imaging capability for highly directive scatterers without sacrificing other scatterers' resolutions. This method takes advantage of the coherence and the phase difference between a pair of 3-D SAR images formed from data collected at two separate circular apertures to eliminate targets that focused at a wrong elevation. In addition, it uses two different transmit frequencies to solve the problem of phase cycle ambiguities. Finally, simulation results validate this new approach.

In the case of high range resolution and squint angle, current chirp scaling algorithm (CSA) and nonlinear CSA (NCSA) have a finite ability to achieve high-quality images. The problem stems from a range-dependent (i.e., space-variant) cubic- and higher order terms of range frequency, which require sufficient compensation or space-variant filtering, in the phase of the synthetic aperture radar transfer function, and this letter aims at dealing with this problem. First, an inequation is introduced to evaluate the highest order of range frequency terms whose space-variant coefficient has to be taken into account. Then, based on the method of series reversion, this letter proposes the extended NCS which can weaken the range dependence of the considered range frequency terms and achieve accurate range cell migration correction and range compression. Simulation results are presented to validate the proposed method.

During the data acquisition of a sliding spotlight or terrain observation by progressive scan (TOPS) synthetic aperture radar (SAR), the steering of the antenna main beam increases the azimuth bandwidth but could result in the azimuth signal aliasing in the Doppler domain. To remove the aliasing, one has used a subaperture method. In this letter, we show a focusing scheme without the use of the subaperture for both sliding spotlight and TOPS SARs. In doing so, we eliminated the obvious increase in data volume or the subaperture division by choosing the pulse repetition frequency that is only 20% greater than the instantaneous bandwidth. The method was incorporated with an available imaging algorithm and then used to process simulated and collected data of the sliding spotlight and TOPS SARs. Well-focused results without aliasing were obtained.

Synthetic aperture radar (SAR) is a major sensor to study internal waves (IWs). In this paper, we numerically simulate IWs in SAR images to investigate the optimal observation conditions of the IWs. In the simulation process, two SAR parameters are considered simultaneously, and different wind fields are also taken into account. Several SAR images from previous experiments (e.g., Surveillance Satellite Project, Synthetic Aperture Radar Internal Wave Signature Experiment, etc.) are referred to validate some of our simulation results. The optimal observation conditions are explored to provide useful guidance to observation and inversion of IWs in SAR images.

Since synthetic aperture technology was employed in radar signal processing, the information capability of radar has greatly been enhanced. A lot of imaging algorithms have also been developed. However, the high-resolution imaging for highly squinted synthetic aperture radar data is still a difficult issue due to large range migration and strong range dependence on the secondary range compression term that is relatively large and cubic with high focusing sensibilities for high resolution. To accommodate for this problem, the "squint-minimization" operation and azimuth nonlinear chirp scaling (CS) (ANCS) operation are studied in this paper. On the basis of these operations, we propose new imaging algorithms and analyze the characteristic of highly squinted data and the difficulty in focusing these data as well as discussing the principle of ANCS. We also introduce a new CS algorithm, and numerical examples show that the proposed algorithm is able to achieve 0.1 m of resolution under a squint angle as large as 70°s.

Recent theory of compressed sensing (CS) suggested that exact recovery of an unknown sparse signal can be achieved from few measurements with overwhelming probability. In this paper, we combine CS technology with a random noise SAR and proposed the concept of random noise SAR based on CS. The block diagram of the radar system and the collected data processing procedure was presented. Theoretic analysis show that the sensing matrix of the random noise SAR exhibits good restricted isometry property (RIP).When the target scene is sparse or sparse in any basis, the random noise radar based on CS can get high accuracy image by collecting far less amount of echo data than traditional noise radar does. The conclusions are all demonstrated by simulation experiments.

This paper analyses ambiguity suppression caused by multiple-input multiple-output (MIMO) SAR azimuth nonuniform samplings. Two methods are analyzed: azimuth spectrum reconstruction algorithm and minimum mean square error (MMSE) imaging algorithm. The azimuth spectrum reconstruction algorithm can reconstruct the scene fine resolution, while the nonideal orthogonality of multi-channel encoding waveforms causes azimuth ambiguous in SAR imaging. The MMSE imaging algorithm can perfectly reconstruct, while it requires high SNR.

In the context of synthetic aperture radar (SAR) image processing and applications, the precise modeling of statistical knowledge is a crucial problem. In this paper, an efficient and flexible statistical model, called generalized Gamma Rayleigh (G¿R) distribution, for amplitude SAR images is proposed by assuming a two-sided generalized Gamma distribution for the real and imaginary parts of the complex SAR backscattered signal. It is shown that the Rayleigh and recently proposed generalized Gaussian Rayleigh distributions can be regarded as special cases of G¿R distribution. Considering that the probability density function estimation problem is formulated as a parameter estimation one for the parametric statistical analysis of SAR images, a two-stage estimator based on second-kind cumulants is derived for the parameters of G¿R distribution. Furthermore, experimental results on several actual SAR images are given to demonstrate the validity and flexibility of the proposed model.

An efficient SAR raw signal simulator accounting for antenna attitude variations is presented here, based on the 2-demensional Fourier domain formulation of SAR raw signal in presence of antenna beam pointing errors. It can meet the requirements of InSAR system simulation to deal with the extended scenes. The validity limit is analyzed to show that this algorithm is suit for the simulation of practical systems.

This paper presented the 3D-SAR with linear array antennas (LAA) which could, in contrast to conventional single-channel 2D-SAR, create the real 3D resolution cells to avoid geometric distortions. Except for conventional side-looking mode, 3D-SAR with LAA can be operated in downward-looking mode which can avoid shadowing effects. The relation between the LAA elements spacing and the elevation angular ambiguity is derived, and the maximal distance between individual antenna elements allowed to avoid elevation angular ambiguity is deduced in this paper. The demonstration of the feasibility of the 3D-SAR with LAA and the relation between elements spacing and elevation angular ambiguity are analyzed by simulation in the last part of this paper.

Aimed at the interferometric calibration problem for the Dual-antenna Airborne InSAR, Considering Ground Control Points(GCPs) are limited, calibration site of area with same height such as flat terrain is advanced. A scheme of establishing parameters bias and building Digital Elevation Model (DEM) is designed. Some airborne InSAR data, derived by Institute Of Electronics, Chinese Academy Of Sciences(IECAS), were used to do calibration experiments with the proposed processor. Their results demonstrated it is efficient.

Ground moving target indication (GMTI) is one of the most important applications in a general bistatic synthetic aperture radar (SAR) system, where the transmitter and receiver move along nonparallel trajectories with different velocities. In order to improve the capability of clutter cancellation in bistatic SAR/GMTI processing, the coherence between two echoes collected by two receivers is investigated, and the full-coherence conditions are derived. A new coherence-improving algorithm for general bistatic SAR complex image pairs is proposed, which can be realized in the following steps: 2-D range azimuth prefiltering processing, relative geometric deformation correction, and image registration. An approximate implementation of 2-D prefiltering and the corresponding prefilter parameter analysis are also given. Last, two numerical experiment results are given to demonstrate the effectiveness of the proposed algorithm.

This paper presented the 3D-SAR based on angle compression principle which could, in contrast to conventional single-channel 2D-SAR, create the real 3D resolution cells to avoid geometric distortions. Except for conventional side-looking mode, 3D-SAR system herein can be operated in downward-looking mode which can avoid shadowing effects. The angle compression principle and the angular ambiguity problem are analyzed in this paper. The analytic expression of angle compression and the condition which should be satisfied to avoid angular ambiguity are also derived in this part. The demonstration of the feasibility of the 3D-SAR based on angle compression principle and the angular ambiguity problem are given by simulation in the last part of this paper.

In this paper, we introduce a new method of terrain slope estimation in the azimuth direction and the ground range direction using only one pass POLSAR data. This method is derived from the polarimetric SAR data compensation for terrain azimuth slopes variation [1] and the Lambertian backscattering model used in the radarclinometry [2]. The AIRSAR L-band POLSAR data are used to show the preliminary results of this method. The comparison of the proposed method and the digital elevation model (DEM) are given to show the effectiveness. The preconditions and limitation of this method are also discussed in this paper.

In this paper, a three-dimensional (3-D) F-SAR imaging algorithm is introduced for Forwarding-looking SAR (F-SAR) data processing. The algorithm is the extension of the two-dimensional (2-D) range stacking algorithm (RSA) and allows the accurate image reconstruction without any interpolation and geometric correction. Then the spatial-varying and spatial-shift properties of the 3-D spread point function (PSF) of Forwarding-looking SAR are revealed with the analytical expression. Finally, the simulation experiment is performed to demonstrate the validity of the 3-D RSA.

Linear array antennas SAR has a resolving capability in the elevation direction, and can get the 3D image of the target. In this paper, we derive the signal model of 3D SAR using a linear array antenna, and get the 3D resolutions and 3D point spread function of array antenna SAR, at the same time the sampling space of array antennas is given. The variance of the resolution in the elevation direction with array antenna angle and referenced look angle is studied. The geometry to reach the best resolution in the elevation direction is analyzed. Meanwhile the resolution for horizontal and vertical antenna array are calculated and compared.

Spatial spectrum estimation technique is applied to resolve layover effect with multi-baseline synthetic aperture radar (SAR) in the paper. Based on the signal model of multi-baseline SAR, the mathematical principle of layover solution with spectrum estimation is derived. The main steps of spatial spectrum estimation technique for layover solution in multi-baseline SAR are obtained. In order to deal with the spectrum ambiguity problem generated by FFT method under limited multi-baseline SAR data, we introduce the Yule-Walker method for spectrum estimation to resolve layover effect. We analyze the principle of Yule-Walker method for performance improvement, and give the processing steps using Yule-Walker method for layover solution in detail. The simulation results for layover solution with FFT method and Yule-Walker method are realized and compared.

Vegetation parameters can be estimated using the single-baseline polarimetric synthetic aperture radar interferometry (POLinSAR) data based on the random volume over ground (RVoG) model. Temporal decorrelation, which is the coherence loss due to scene changes within the time between radar data acquisitions, will decrease the estimation accuracy and needs to be compensated. The RVoG+VTD model is a simple model incorporating a temporal decorrelation term into the RVoG model. The inversion of RVoG+VTD model can not perform due to the limited number of single-baseline POLinSAR observables. Dual-baseline POLinSAR approach provides more observables and hence can be used to invert the model. This paper introduces and analyzes the dual-baseline inversion procedure of RVoG+VTD model and validates them using simulated data.

An estimation of the number of clusters is proposed for fully polarimetric SAR data analysis, and a corresponding unsupervised segmentation algorithm is also given based on the Cloude-Pottier decomposition and the complex Wishart clustering. The Monte-Carlo Cross-Validation (MCCV) is used to estimate the optimal number of clusters to reveal the inner structure of the data. Since it is a quantitative estimation of the classification performance, the MCCV algorithm also has the potential capability to perform the unsupervised segmentation validation. The effectiveness of the MCCV estimation and the segmentation algorithm is demonstrated using ESAR data acquired.

In this paper, the cross validation algorithm is used to estimate the number of clusters for the unsupervised classification of fully polarimetric SAR data. Three different cross validation algorithms are applied for comparison, which are the dispersion measure method, the V-fold cross validation (VFCV) and the Monte-Carlo cross validation (MCCV). Our current experiments show that the dispersion measure method appears generally unable to provide a reliable estimation. The VFCV and the MCCV algorithms seem to be more effective than the dispersion measure method. Moreover, the VFCV is much faster than the MCCV, but the MCCV may be able to provide better estimation than the VFCV.

In this paper, first compared different receive channel schemes for InSAR application, then the main principal and technique of three practical internal calibration schemes are analyzed and compared, especially the famous SIR-C and X-SAR internal calibration system. Based on these different schemes, a new internal calibration scheme using microwave over fiber link is subscribed.

In this letter, a texture-preserving despeckling algorithm for synthetic aperture radar images using an evidence framework is proposed. The salient aspects of this approach are given as follows. (1) The maximuma posterioriestimate can be guaranteed to converge to the optima by selecting the Gaussian distribution and Gaussian Markov random field model as the likelihood function and prior model, respectively. (2) MacKay's evidence framework can automatically sustain the balance between speckle reduction and texture preservation. (3) We use the Jeffreys prior to perform the second-level inference of the evidence framework. Experimental results are given to demonstrate the validity of the proposed despeckling method.

In this paper, an unsupervised segmentation is proposed for fully polarimetric synthetic aperture radar (SAR) data analysis. The backscattering powerSPANcombined withH/alpha/Ais used to obtain the initial cluster centers. We use the Wishart test statistic to perform an agglomerative hierarchical clustering to obtain the segmentation results with different numbers of clusters. The appropriate number of clusters is automatically estimated using the data log-likelihood (Lm), and the resulting images with the estimated number of clusters are the final segmentation results. The experiments show that theSPANhas additional information that is not contained inH/alpha/A, and this information could be useful for the initialization. The number of clusters seems to be a crucial point for the segmentation, which will affect the segmentation performance. It is also shown that the data log-likelihood has the potential ability to reveal the inner structure of fully polarimetric SAR data.

In this paper, the IHSL transform and the fuzzy C-means (FCM) segmentation algorithm are combined together to perform the unsupervised classification for fully polarimetric SAR data. We apply the IHSL colour transform to H/alpha/SPAN space to obtain a new space (RGB colour space) which has a uniform distinguishability among inner parameters and contains the whole polarimetric information in H/alpha/SPAN. Then the fuzzy C-means algorithm is applied to this RGB space to finish the classification procedure. The main advantages of this method are that the parameters in the color space have similar interclass distinguishability, thus it can achieve a high performance in the pixel based segmentation algorithm, and since we can treat the parameters in the same way, the segmentation procedure can be simplified. The experiments show that it can provide an improved classification result compared with the method which uses the H/alpha/SPAN space directly during the segmentation procedure.

In this paper, a new nonlinear magnification method is proposed. The new method produces the effect of nonlinear magnification based on perspective projection and the Bezier curve is used as drop-off function. So the new method can enhance the local information and keep the global context. It can provide different representation by adjusting the distortion degree of nonlinear magnification especially. In this way, image interpretation and target recognition would be performed effectively.

In this paper a new method is proposed to solve the problem of high resolution satellite imagery without GCPs by considering the consecutive imaging parameters. That is, the imaging parameters of consecutive imagery are calculated based on GCPs in order to set up the prediction formula, and then the imaging parameters of high resolution imagery can be forecasted. Thus the rigorous model is introduced to precisely locate imagery. QuickBird imagery is test and geo-referencing accuracy reaches 3-4 pixels.

Along with the great advancement of remote sensing technology and extensive application of remote sensing information, more and more requirements on capability and efficiency for all kinds of interpreting system of remote sending images come out. The conventional interpreting systems commonly have the operating mode with one single interpreter and an isolated computer. But this conventional mode has a serious immanent limitation due to the isolation of position, personnel, data, software, hardware and so on. This paper provides a CSCW-based interpreting system of remote sensing imagery (RSI) to avoid the limitation from the conventional systems. This new system fully integrates the remote sensing technology and computer supported cooperative work (CSCW) technology, actualizes interpreting of RSI with a synchronized operation for multi-user. Consequently it greatly improves capability, efficiency, and accuracy. The better actual effect proves that interpreting system of RSI is feasible and effective.

In this paper, a matching method based on feature points matching is proposed. First, a point detector is used to detect interest points of a source image and a target image. Then genetic algorithms (GAs), which are useful in finding global optima, are used to match the detected points. A fitness function which is proven rotation-invariance is proposed to measure the similarity of two points in the paper. According to the results of the GAs, the corresponding points of the target image are found in the source image. The algorithm can overcome the rotation distortion of the images. The experiment results confirm the proposed algorithm is efficient and robust.

SAR/ISAR image processing involves a 2D Fourier transform that produces high intensity sidelobes which obscure low intensity scatters in the image. Although the sidelobe can be reduced using parametric windows, the image resolution becomes worse. The apodization technique can reduce sidelobe level while maintaining the image resolution. In this paper, based on the analysis of the apodization algorithm, a modified apodization method was present via stronger constraint. The modified method has higher resolution and lower sidelobe level than the original method. This method was proved correct with a real ISAR image.

In this paper, we introduce a new method to locate the remote images. By using the Geographic Information System(GIS) resources, we can extract the coastlines information. At the same time, by using image processing technologies, we can also extract the similar information from remote images. Based on the image matching technologies, we can roughly locate the remote images. Some simulation experiments are implemented on remote sensing images with coastlines. Preliminary results verify the feasibility of the technique.