Structural load suppression of wind turbines is one of the important means to improve the economic efficiency of wind farm operation. However, wind turbines are complex systems coupled with multiple control variables. There is no comprehensive solution for the flexible and coordinated control of multiple controllable degrees of freedom. The main difficulty lies in the greater optimization burden of multi-degree-of-freedom systems. This paper proposes a variable degree-of-freedom control system for wind turbines, and uses multivariable model predictive control to design collaborative controllers for generator torque, pitch angle and yaw angle. To avoid the increase in complexity caused by the continuous intervention of yaw control, a yaw control cut-in/cut-out strategy is set, and the non-disturbance of switching is theoretically proved. Simulation results show that the proposed variable degree-of-freedom control system can effectively realize active power control and structural load suppression of wind turbines in a large wind speed range.

Land cover changes (LCCs) affect surface temperatures at local scale through biophysical processes. However, previous studies on the temperature effects of LCCs, whether the potential impacts of virtual LCCs using the space-for-time assumption or the actual impacts of observed LCCs using the space-and-time scheme, have primarily concentrated on analyzing their spatial distribution patterns. Consequently, the temporal trends of temperature effects due to LCCs are less discussed. This study analyzed the temporal trends of land surface temperature (LST) effects induced by actual LCCs by using long-term European Space Agency land cover data and Advanced Very High Resolution Radiometer LST data. The results show that, from 2001 to 2016, there was a gradual reduction in the count of pixels experiencing LCCs globally in which cultivated land expansion is an important cause of LCC. The LST's response to actual LCCs presented a trend of initial increase followed by a subsequent decrease.

The classification of dominant leaf types, which distinguishes forests based on their leaf conditions, is beneficial for forest management and policymakers. This paper proposes a model based on U-Net to classify the land into non-tree areas, broadleaf forests, and coniferous forests. The dual-pol Sentinel-1 data from January, May, August, and October of 2018 were stacked as a time series. Due to the class imbalance issue, where the non-tree area category dominates (52.69%) the dataset, the model tends to be biased. Thus, re-weighting is introduced to balance the loss. We tested and compared two types of methods: class-aware and task-aware re-weighting. The results indicate that re-weighting effectively mitigates the class imbalance issue.

This study explores the application of resampling techniques to address the data imbalance issue in LiDAR point cloud-based tree species classification. Considering the data imbalance problem in the point clouds, a simple yet efficient method, resampling is introduced in this paper. We compared two resampling strategies, undersampling the majority classes and oversampling the minority classes. The experimental results show that deep learning models, particularly when augmented with resampling strategies, can significantly improve the classification accuracy. Both strategies increased the overall accuracy (by 0.8% and 2% respectively) and classification of the minority class by 5.56%. And oversampling is superior to undersampling because it makes use of all the training samples.

Leaf area index (LAI) is a pivotal biophysical parameter for characterizing canopy structure and monitoring vegetation growth. Although the look-up table (LUT) method has been widely employed for LAI retrieval, the optimization of key retrieval processes remains to be explored. Here, we proposed a generic optimization strategy for LUT-based inversion based on Landsat -8 imagery and global ground LAI measurements. Specifically, based on the LUT generated by the PROSAIL model, LAI inversion was optimized by introducing several functions, including band selection, artificial noise addition, cost function (CF) substitution, and multiple solutions. Furthermore, the optimized LUT-based inversion method was compared to the Simplified Level 2 Product Prototype Processor (SL2P) method and the ground-measurement-derived (GMD) regression method to comprehensively evaluate its performance over various vegetation types. Results showed that the combination of Red, near-infrared (NIR), and shortwave infrared-1 (SWIR1) bands was well suited to capture LAI dynamics. In terms of accuracy and efficiency, the best performance was achieved by the optimal band combination and retrieval parameter settings (i.e., root-mean-square error (RMSE) as CF, noise level of 20%, and multiple solutions of 5%), with the RMSE and${R} ^{2}$of 0.817 and 0.740, respectively. In addition, the optimized LUT-based inversion was superior to SL2P method in accuracy and to GMD regression method in efficiency. Overall, the optimized LUT-based inversion strategy can be applied for estimating decametric-resolution LAI with high accuracy over different regions and observation dates at a global scale, exhibiting high adaptability and generalization capability, especially for crops, and requiring no ground LAI measurements.

Recently deep learning models were applied to detect fast short-term volcanic deformations using interferometric synthetic aperture radar (InSAR) data. However, volcanic deformation detection is limited by the availability of real positive samples. In previous work, we used hybrid synthetic-real InSAR deformation maps set to train an InceptionResNet v2 model capable of detecting deformations down to 5 mm/year in real set. However, our model also reported false positive detections. One possible reason is the data distribution gap between the real and hybrid sets. In this paper, an experiment is conducted to analyze the gap between the hybrid and real sets that resulted in false positives. Three subsets of the fine-tuning set are created based on t-SNE analysis using different sampling strategies. The classification model is fine-tuned using these subsets. The results show that the strategy of removing only the most confusing examples and keeping the larger data set size reduces the false positive rate from 32.29% to 27.01%.

High quality training samples are essential for crop mapping. However, since traditional sample acquisition methods are based on expert interpretation or field research, they are time-consuming and expensive. Using the unique time window of a crop, it is possible to distinguish a specific crop from other features. Therefore, using phenological information combined with machine learning methods for sample migration is a very feasible solution for crop mapping. In this study, we developed a yearly automated generated sample migration algorithm based on crop phenological features. Using image time series derived from data acquired by Landsat sensor systems 5, 7, 8 accessible through the Google Earth Engine cloud data platform, we developed a procedure for temporally displacing ground-truth soybean samples based on phenological features of the crop. With these data, we then generated annual maps of soybean in Heilongjiang Province, China. Overall accuracy of the temporally displaced soybean samples was higher than 95%, while the overall accuracy of the soybean maps obtained was more than 83%. This study provides a feasible approach for developing ground-truth samples from long term image time series, suitable for mapping the dynamics of crops across space and time.

In this paper, we proposed a new framework for 3D point cloud simulation for forest above-ground biomass estimation. It takes tree variables as input and automatically generates 30m by 30m scenes and simulates their corresponding Li-DAR point clouds. 2000 3D tree models of 10 species are generated, with which 5000 forest scenes representing four types of eco-regions are built. Their corresponding biomass are then calculated with allometric equations. We used the simulated tropical scenes (1000 samples) to test four classical machine learning models’ ability in biomass estimation from point clouds. Experimental results show that data augmentation is able to significantly boost test accuracy; self-supervised learning can improve the estimation results; among the four models, ResNet18 is the best baseline model which has achieved a R-square score of 0.6615, and reduced the root mean square error to 202 kg (mean biomass in our dataset is 2114kg).

LiDAR scanning data, which is able to acquire the vertical structures of forests, is of great potential in forest monitoring and biodiversity quantification. Besides, the derivation of some forest indices, such as biomass, relies on individual tree delineation (ITD). In this paper, we generated a dataset for individual tree delineation using LiDAR-derived point clouds. This dataset can be used to fairly compare different ITD methods and to develop deep learning algorithms for tree segmentation. The acquired LiDAR data consist of 0.94 billion points covering an area of about 31 km2in the Netherlands. We first used a rule-based algorithm to remove non-tree points. And then a mean shift clustering method is utilized to segment the points. Besides, we proposed a method that compares the highest point in the same cluster to evaluate the delineation results. In the future, the derived segmentation result will be compared with existing individual tree delineation algorithms.

Subtle volcanic deformations point to volcanic activities, and monitoring them helps predict eruptions. Today, it is possible to remotely detect volcanic deformation in mm/year scale thanks to advances in interferometric Synthetic Aperture Radar (InSAR). This paper proposes a framework based on a deep learning model to automatically discriminate subtle volcanic deformations from other deformation types in five-year-long InSAR stacks. Models are trained on a synthetic training set. To better understand and improve the models, explainable AI analyses are performed. In initial models, gradient-weighted Class Activation Mapping (Grad-CAM) linked new-found patterns of slope processes and salt lake deformations to false-positive detections. The models are then improved by fine-tuning with a hybrid synthetic-real data, and additional performance is extracted by low-pass spatial filtering of the real test set. T-SNE latent feature visualization confirmed the similarity and shortcomings of the fine-tuning set, highlighting the problem of elevation components in residual tropospheric noise. After fine-tuning, all the volcanic deformations are detected, including the smallest one, Lazufre, deforming 5 mm/year. The first time confirmed deformation of Cerro El Condor is observed, deforming 9.9-17.5 mm/year. Finally, sensitivity analysis uncovered the model’s minimal detectable deformation of 2 mm/year.

With improvement in the processing of synthetic aperture radar interferometry (InSAR) data, the detection of long-term volcanic deformations becomes possible. While deep learning (DL) models are considered black-box models, challenging to debug, the advances in explainable AI (XAI) help understand the model and how it makes decisions. In this paper, the model is trained on synthetic InSAR velocity maps to detect slow, sustained deformations. XAI tools, including Grad-CAM and t-SNE, are utilized for understanding and improving the trained model. Grad-CAM helps identify the slope-induced signal and salt lake patterns responsible for the model’s mis-classifications. T-SNE feature representation visualizations are used to estimate data sets and model class separation ability. Additionally, a sensitivity analysis shows the model performance with different intensity deformation data and uncovers the minimal detectable deformations of 1 cm cumulative deformation over five years.

In order to estimate tree biomass, allometric equations take tree parameters such as tree height, wood density, circumference of trunk, and crown diameter as input parameters. Given that most of these quantities are challenging to be extracted from remote sensing data, we evaluate the option to approximate biomass by tree height only. We study our approach by evaluating linear regression, random forest, and Gaussian process regressor models when applied to the 2016 Jucker dataset. Results indicate that linear models fail to properly capture the relationship between biomass and tree height, but the Gaussian process regressor outperms the other two candidate models.

Earth observation data has huge potential to enrich our knowledge about our planet. An important step in many Earth observation tasks is semantic segmentation. Generally, a large number of pixelwise labeled images are required to train deep models for supervised semantic segmentation. On the contrary, strong inter-sensor and geographic variations impede the availability of annotated training data in Earth observation. In practice, most Earth observation tasks use only the target scene without assuming availability of any additional scene, labeled or unlabeled. Keeping in mind such constraints, we propose a semantic segmentation method that learns to segment from a single scene, without using any annotation. Earth observation scenes are generally larger than those encountered in typical computer vision datasets. Exploiting this, the proposed method samples smaller unlabeled patches from the scene. For each patch an alternate view is generated by simple transformations, e.g., addition of noise. Both views are then processed through a two-stream network and weights are iteratively refined using deep clustering, spatial consistency, and contrastive learning in the pixel space. The proposed model automatically segregates the major classes present in the scene and produces the segmentation map. Extensive experiments on four Earth observation datasets collected by different sensors show the effectiveness of the proposed method. Implementation is available at https://gitlab.lrz.de/ai4eo/cd/-/tree/main/unsupContrastiveSemanticSeg.

Deep learning-based synthetic aperture radar (SAR) target recognition often suffers from sparsely distributed training samples and rapid angular variations due to scattering scintillation. Thus, data-driven SAR target recognition is considered a typical few-shot learning (FSL) task. This article first reviews the key issues of FSL and provides a definition of the FSL task. A novel adversarial autoencoder (AAE) is then proposed as an SAR representation and generation network. It consists of a generator network that decodes target knowledge to SAR images and an adversarial discriminator network that not only learns to discriminate “fake” generated images from real ones but also encodes the input SAR image back to target knowledge. The discriminator employs progressively expanding convolution layers and a corresponding layer-by-layer training strategy. It uses two cyclic loss functions to enforce consistency between the inputs and outputs. Moreover, rotated cropping is introduced as a mechanism to address the challenge of representing the target orientation. The moving and stationary Target recognition (MSTAR) 7-target dataset is used to evaluate the AAE’s performance, and the results demonstrate its ability to generate SAR images with aspect angular diversity. Using only 90 training samples with at least 25° of orientation interval, the trained AAE is able to generate the remaining 1748 samples of other orientation angles with an unprecedented level of fidelity. Thus, it can be used for data augmentation in SAR target recognition FSL tasks. Our experimental results show that the AAE could boost the test accuracy by 5.77%.

Due to the small size of foreign object debris (FOD) and varied complex weather conditions, the detection of FOD on airport runways is a great challenge. Radar is an important method for detecting FOD targets. However, almost all the existing systems are based on real apertures, which have disadvantages such as low azimuth resolution and susceptibility to rain interference. Here, an innovative FOD detection radar system based on arc-scanning synthetic aperture radar (AS-SAR) technology, the AS-SAR based FOD detection system (AS-FODR), achieves omnidirectional coverage with a very high azimuth resolution and the suppression of flicker clutter, such as rain drops in severe weather. According to the radar imaging simulation of a scene under rainy conditions and the information processing analysis of field experiments, a prototype system was built, and an efficient data process flow was proposed. In short, a one centimeter FOD target was detected on the runway more than 250 m away, proving that the use of AS-FODR is feasible and effective.

The CCCTC-binding factor (CTCF) mediates transcriptional regulation and implicates epigenetic modifications in cancers. However, the systematically unveiling inverse regulatory relationship between CTCF and epigenetic modifications still remains unclear, especially the mechanism by which histone modification mediates CTCF binding. Here, we developed a systematic approach to investigate how epigenetic changes affect CTCF binding. Through integration analysis of CTCF binding in 30 cell lines, we concluded that CTCF generally binds with higher intensity in normal cell lines than that in cancers, and higher intensity in genome regions closed to transcription start sites. To facilitate the better understanding of their associations, we constructed linear mixed-effect models to analyze the effects of the epigenetic modifications on CTCF binding in four cancer cell lines and six normal cell lines, and identified seven epigenetic modifications as potential epigenetic patterns that influence CTCF binding intensity in promoter regions and six epigenetic modifications in enhancer regions. Further analysis of the effects in different locations revealed that the epigenetic regulation of CTCF binding was location-specific and cancer cell line-specific. Moreover, H3K4me2 and H3K9ac showed the potential association with immune regulation of disease. Taken together, our method can contribute to improve the understanding of the epigenetic regulation of CTCF binding and provide potential therapeutic targets for treating tumors associated with CTCF.

Drug repositioning has become a prevailing tactic as this strategy is efficient, economical and low risk for drug discovery. Meanwhile, recent studies have confirmed that small-molecule drugs can modulate the expression of disease-related miRNAs, which indicates that miRNAs are promising therapeutic targets for complex diseases. In this study, we put forward and verified the hypothesis that drugs with similar miRNA profiles may share similar therapeutic properties. Furthermore, a comprehensive drug-drug interaction network was constructed based on curated drug-miRNA associations. Through random network comparison, topological structure analysis and network module extraction, we found that the closely linked drugs in the network tend to treat the same diseases. Additionally, the curated drug-disease relationships (from the CTD) and random walk with restarts algorithm were utilized on the drug-drug interaction network to identify the potential drugs for a given disease. Both internal validation (leave-one-out cross-validation) and external validation (independent drug-disease data set from the ChEMBL) demonstrated the effectiveness of the proposed approach. Finally, by integrating drug-miRNA and miRNA-disease information, we also explain the modes of action of drugs in the view of miRNA regulation. In summary, our work could determine novel and credible drug indications and offer novel insights and valuable perspectives for drug repositioning.

A zero-shot learning (ZSL) method of automatic target recognition (ATR) in synthetic aperture radar (SAR) image is proposed to address the scenario, where no SAR sample of a particular target is available for training. To learn features of the unseen target, physics-based electromagnetic (EM) simulated images of the target under different azimuth angles are used as the training data instead. The challenge lies in the fact that the simulated image has a distinct but nonessential texture that the real images do not have and, thus, can easily result in an overfitted discriminator network. To overcome this problem, all images are first preprocessed with a nonessential factor suppression step and then fed into a pretrained convolutional neural network for feature extraction. Finally, the feature vector is fed into a trainable fully-connected network for classification. The low-dimensional embedding of feature vectors suggests that the nonessential factor suppression can align the simulated samples with true samples effectively. We propose the max-tolerability principle and averaged margin index for ZSL, which is a useful indicator for selecting optimal classifier. We validated our method on ten-type target recognition task on MSTAR data sets and achieved 91.93% accuracy on nine known targets and 79.08% accuracy on zero-shot target.

This paper focuses on the generalization ability of model for SAR automatic target recognition (ATR). An object-based similarity evaluation method for MSTAR datasets is proposed at first to show the relationship between classification accuracy and orientation difference between training and test images. It reveals poor orientation generalization ability of traditional methods for orientation interval larger than 10deg. In order to improve the orientation generalization ability, a novel adversarial autoencoder neural networks (AAN) is proposed in this paper. It learns a code-image-code cyclic network by adversarial training for the purpose of generating new samples at different azimuth angles. The learned orientation predictor and classifier is applied to test samples. Proposed network achieved over 86% classification accuracy on 7-type MSTAR datasets when minimum orientation interval is limited to 25deg, and is about 4% higher than baseline model A-ConvNets under the same condition.

We propose a deep neural networks based method to reconstruct full polarimetric (full-pol) information from single polarimetric (single-pol) SAR data. It consists of two parts: feature extractor which is used to obtain multi-scale multi-layer features of targets in single-pol gray image, and feature translator that converts the geometric features to defined polarimetric feature space. The proposed method is demonstrated on L-band UAVSAR of NASA/JPL images over San Diego, CA, and New Orleans LA, USA. Both qualitative and quantitative results show the reconstructed full-pol images agree well with true full-pol images, the proposed networks have a good spatial robustness. Model-based target decomposition and unsupervised classification can be used directly on constructed full-pol images.

Artificial intelligence such as deep learning has become the dominant approach in computer vision area. It has great potential in improving the performance of SAR automatic target recognition (ATR) as well. In this paper, we present a framework for intelligence SAR ship recognition and a preliminary implementation as well as a demonstration with the ALOS2 data.

SUMMARY: As a promising field of individualized therapy, non-coding RNA pharmacogenomics promotes the understanding of different individual responses to certain drugs and acts as a reasonable reference for clinical treatment. However, relevant information is scattered across the published literature, which is inconvenient for researchers to explore non-coding RNAs that are involved in drug resistance. To address this, we systemically identified validated and predicted drug resistance-associated microRNAs and long non-coding RNAs through manual curation and computational analysis. Subsequently, we constructed an omnibus repository named ncDR, which furnishes a user-friendly interface that allows for convenient browsing, visualization, querying and downloading of data. Given the rapidly increasing interest in precision medicine, ncDR will significantly improve our understanding of the roles of regulatory non-coding RNAs in drug resistance and has the potential to be a timely and valuable resource. AVAILABILITY AND IMPLEMENTATION: http://www.jianglab.cn/ncDR/. CONTACT: [email protected] or [email protected]. SUPPLEMENTARY INFORMATION: Supplementary data are available at Bioinformatics online.

As is well known, multibaseline phase unwrapping (PU) is put forward to overcome single-baseline PU in discontinuous-terrain-height estimation. This letter presents a refined algorithm based on the cluster analysis (CA)-based noise-robust efficient multibaseline PU algorithm proposed by H. Yu. The basic idea is to combine multiple interferometric synthetic aperture radar interferograms with different baseline lengths by a linear combination. The new interferograms after the linear combination increase the ambiguity heights. The number of resulting groups on the envelope of the intercept histogram is decreased and the distance between different intercept groups is widened. Compared with the conventional CA method, the significant advantage of the refined CA (RCA) algorithm is that it improves noise robustness when the intercept groups are densely distributed. The proposed RCA algorithm is validated using the simulated interferometric data. The results demonstrate that the noise robustness performance is better than that of the CA method.

Zero-shot learning (ZSL) is of critical importance for practical synthetic aperture radar (SAR) automatic target recognition (ATR) as training samples are not always available for all targets and all observation configurations. We propose a novel generative-based deep neural network framework for ZSL of SAR ATR. The key component of the framework is a generative deconvolutional neural network referred to as generator. It learns a faithful hierarchical representation of known targets while automatically constructing a continuous SAR target feature space spanned by orientation-invariant features and orientation angle. It is then used as a reference to design and initialize an interpreter convolutional neural network, which is inversely symmetric to the generator network. The interpreter network is then trained to map any input SAR image, including those of unseen targets, into the target feature space. In a preliminary experiment with the Moving and Stationary Target Acquisition and Recognition data set, seven targets are used in the training of generator and interpreter networks. Then, the eighth target is used to test the interpreter, where it is correctly mapped to the reasonable spot spanned by the previous seven targets and its orientation can also be estimated.

This paper reformulates the problem of polarimetric incoherent target decomposition as a general image factorization which aims to simultaneously estimate a dictionary of meaningful atom scatterers and their corresponding spatial distribution maps. Both model-based and eigenanalysis-based decompositions can be seen as special cases of image factorization under specific constraints. The inverse problem of image factorization can be converted to an equivalent nonnegative matrix factorization (NMF) problem via redundant coding. It enables a wide range of NMF algorithms with various regularizations to be directly applicable to polarimetric image analysis. The advantage of the proposed image factorization is demonstrated on both synthesized and real data. It also shows that extended applications such as speckle reduction and classification can benefit from the proposed image factorization.

The interferometric radar which is used to monitor and measure the displacement of large artificial buildings is one of the new advanced technology in recent years, but the research on interferometric radar about the comparison of radar signal system and the argument about the key parameters of radar are rarely seen. Firstly, the basic principle of interferometry was introduced in this paper, then the general constraint relation between the no fuzzy deformation measurement and the data update rate of radar was induced. After then, this general constraint relation was used in analyzing radar signal system and designing experiments, proved the importance of the constraint relation in the interferometric radar of displacement measurement.

Radar sensors have received more and more interest for unmanned ground vehicle to sense positive and negative obstacles in unstructured environments or out fields, especially on negative obstacle. In this paper, we present an approach for extracting the features of obstacles from radar images. Based on interferometric synthetic aperture radar (InSAR) images focused by the back-projection (BP) algorithm, range compensation, speckle filtering and threshold segmentation are performed. And morphological operations are used to perform some simple connectivity filtering to smooth the image and remove spurious pixels. Finally, feature fusion is applied to the amplitude and the correlation coefficient images. Both the theoretical analysis and the experimental results indicate that the proposed method is an efficient method.

Off-road positive and negative obstacle detection is a challenge problem to be solved by unmanned ground vehicle. Traditional sensors, such as: optical camera, lidar and millimeter wave radar, have limited performance in off-road environments, especially when obstacles are far away or covered by sparse grasses. We have proposed a forward-looking InSAR sensor to tackle the problem and have built a rail-based InSAR prototype. The forward-looking InSAR can provide more information of harsh off-road environments than existing unmanned ground vehicle (UGV) based radars. The forward-looking InSAR can provide a scattering image, a coherence image and a digital terrain model (DTM) of the same scene ahead the radar during each scan. Each type of image can highlight some unique features of an obstacle. In this paper, an obstacle detection method is proposed by combining the shadow feature and the edge scattering feature. The principle method is close related to the scattering property difference between positive obstacles, negative obstacles and other objects. Positive obstacle feature large amplitude followed by low coherence area; while at the same time, negative obstacles feature low coherence area followed by large amplitude. Other objects don't have the unique feature. To mitigate false alarms, shadows are segmented in coherence images, and edge scattering features are extracted in scattering image. Firstly, the coherence image is converted into a binary image by applying a threshold. Shadow areas are roughly segmented as their coherences are low. Then the binary image is filtered by morphologic opening operation to eliminate small patches. Subsequently, an edge detection operation is applied to the filtered image. The edges of positive and negative obstacle are among the detected edge image. For each position of the detected edge, a cut is performed on the same position in the scattering image to extract a slice along the range direction. The judgment is formed by calculating the energy ratio between the near half slice of the farther half slice. Finally, positive and negative obstacles can be discriminated by comparing the judgment with two thresholds in an unsupervised fashion. We have conducted a field experiment on a ground covered by sparse grasses. A pit and a mound are deliberately built in the experiment scene. Experimental results have validated the proposed method.

This paper proposes a new formulation for POLSAR incoherent target decomposition, which models the POLSAR image as a result of a two-stage process, i.e. assembly of coherent scatterers followed by spatial distribution of incoherent atom scatterers. This assembly-distribution forward model is described by a triple matrix multiplication while its inverse problem, the new target decomposition problem, can be solved via nonnegative matrix factorization. The proposed new target decomposition has similarity to image clustering. Some preliminary results are presented.

In foot-mounted positioning systems, it is hard to align multi-agent trajectories. In addition, the positioning accuracy is hard to maintain due to inertial drifts. An approach for trajectory initialization and calibration using iBeacons is proposed in this paper. This approach is under the framework of a particle filter. In the observation model of the particle filter, a nonparametric Gaussian Process (GP) regression model is adopted to describe the relationship between the estimated range and the observed RSS. Then the weights of the particles are updated according to the trained GP. GP is adopted here because it not only considers the sensor noise, but also the uncertainty in the model, which denotes the multi-path effects, human sheltering effects and so on in receiving the iBeacon signals. At last, a large-scale real-scenario experiment is carried out with a total walking length of about 5.4 kilometers. The results have demonstrated the effectiveness of the proposed approach for trajectory initialization and calibration, with the final positioning error reduced from 85.4 meters to only less than 1meter.

Foot-mounted inertial sensor is a practical and efficient approach for indoor personnel positioning, because it requires neither pre-deployed infrastructure nor on-line “fingerprint” database. To suppress the heading drift of gyroscopes, magnetic field is often used as a stable orientation reference. But the magnetic field inside buildings changes rapidly while space and time changes, and it is also sensitive to some metal objects or electrical equipments. Thanks to the regular structures of most buildings, Heuristic Drift Elimination (HDE) techniques can be applied to suppress slowing-changed small-scale drifts between two steps, under the assumption that most of the paths in the buildings consist of straight lines and fixed angles. Inspired by the heuristic method, a Near Real-Time Heading Drift Correction (NRHDC) approach is proposed in this paper, which determines the current walking behavior based on the steps foregoing and following. Different step status are set to classify each step and different templates are trained for walking behavior discrimination. This approach is supposed to exploit more information and give more precise result. The experimental results in a building also demonstrate the effectiveness of the proposed approach, with superior accuracy to other methods.

The Zero-velocity Update (ZUPT)-aided Extended Kalman Filter(EKF) algorithm is commonly used to suppress the error growth of the inertial based pedestrian navigation systems, but it still suffers from long-term heading drift. The magnetic field was suggested to mitigate the heading errors of for positioning and navigation, but it undergoes severe perturbation in indoor scenarios. The Quasi-Static magnetic Field (QSF) method was developed to estimate heading errors using magnetic field in perturbed environments. However, this method may bring extra errors to system because of the high false alarm probability of detecting the quasi-static field. In this paper, we propose a heading error estimation approach based on the improved QSF approach for foot-mounted Inertial Navigation Systems (INS). For the approach, a magnetometer calibration method is developed to eliminate the deviation caused by the positioning system platform and shoes firstly. Then an improved QSF detection approach is proposed and then used for generating the desired magnetic measurements which are fed into the EKF to estimate the heading errors. The experiments indicate that the proposed method is effective in suppressing the heading errors.

To tackle the error accumulation problems of the foot-mounted inertial positioning systems, aiding measurements are often adopted to calibrate the inertial-generated raw trajectories. Currently, the calibration methods, although effective, still lack interior connection and unity. In this paper, a unified anchor based trajectory calibration framework is proposed. The concept of anchors is defined as different types of aiding information and is considered as different constraint conditions to the inertial-generated trajectory. The foundation of the framework is a particle filter, and different anchor information correspond to different particle weight updating strategy. The proposed framework is expansive to incorporated different types of anchors. We choose three representative anchors to elaborate the proposed framework and they are: Ultra-WideBand (UWB) ranging anchors, iBeacons and the building structure based virtual anchors. The results in simulations and real scenario tests have validated the effectiveness of the proposed anchor based framework in calibrating the inertial-generated raw trajectories.

In this letter, we investigate an application of a generalized polar decomposition of the Mueller matrix for polarimetric synthetic aperture radar (Pol-SAR) classification. Six roll-invariant parameters (diattenuation, retardance, polarization power, depolarization anisotropy, depolarization power, and transmittance) are selected as features for the classification of scattering types. Experimental results using the AIRSAR data over Flevoland show that, for most field types, the D-R-AΔ- Δ-m00set provides the highest classification accuracy, followed by the D-R-PΔ-Δ-m00set which also provides better accuracy than the widely used H -α-(δ-γ)-A-span set. The proposed method would be valuable for Pol-SAR interpretation.

The pairwise separability index (SI) has been demonstrated as an effective indicator for capturing crucial phenological differences between two plant species. However, its application to crop types, which have more obvious phenological characteristics than natural vegetation, has received less attention, and extending the pairwise SI to multiple crops for feature selection still remains a challenge. This paper presented two SI extension approaches (SIaveand SImin) to select the optimal spectro-temporal features for multiple crops, and investigated their classification performance using Heilongjiang Province, China, as a study area. Feature interpretability and classification accuracy of different crops were evaluated for the two approaches. The results showed that the SIaveapproach generally has relatively high feature interpretability due to its better description of crucial phenological characteristics of different crops. Those crops with high separability are insensitive to the extension approach and have similar classification accuracy for the two approaches, whereas those crops with poor separability show good performance with the SIminmethod. Due to the higher temporal autocorrelation, the optimal features for crop classification that are selected by the SIaveapproach exhibit greater information redundancy across the time domain than those that are selected by the SIminapproach, which largely explains the relatively low classification accuracy achieved using the SIaveapproach. These comparison results between SIminand SIaveapproaches also indicate that time-series images with high temporal resolution do not necessarily produce high classification accuracy, regardless of their ability to describe the seasonal characteristics of crops.

An applicable inertial-based pedestrian navigation system (PNS) is often composed of two stages: trajectory-generation stage and trajectory-calibration stage. As a mature algorithm, ZUPT (Zero-velocity Update)-aided EKF (Extended Kalman Filter) is commonly deployed to resolve trajectories of pedestrians, with short-term drifts suppressed during stance phase. In the trajectory-calibration stage, many a priori knowledge based methods are adopted to suppress or even eliminate long-term drifts. In this paper, we propose a particle filter based approach for trajectory calibration with awareness of the rectangular structures of buildings. The navigation frame is divided into eight directions, including four “domain” directions and four complementary directions. The “domain” directions are consistent with the layout of the rectangular building structure, while the complementary directions are set to avoid abrupt adjustment in minority situations, such as turnings and random walks. In the particle filter framework, the headings and positions from the previous stage are adjusted by assigning weight to particles according to a Gaussian function, which over-weight particles nearby the eight directions and de-weight those more far away from them. Along with the resampling procedure, long-term heading drifts are suppressed by gradually eliminating particles at odds with the rectangular building structure. To verify the accuracy and robustness of the proposed approach, the authors have conducted many real-world experiments in different scenarios. In a typical office building experiment with corridors and stairs, the location error is less than 1% of total walking distance, which is acceptable in most applications. Another relatively large-scale walk in a mall with curves and lines demonstrates the adaptability of our approach to some special situations with complex paths. The results have shown that our approach can perform accurate, continuous and stable positioning.

For Search and Rescue (SAR) applications in large buildings, vertical accuracy may seem to be more important than horizontal accuracy, because correctness of floor estimation is mission critical. In this paper, an approachfor height error suppression based on apriori knowledge is proposed, which focuses on long-term height drift suppression along with proper floor determination. An Extended Kalman Filter (EKF) algorithm is firstly reviewed, which integrates barometer data to correct height drift brought by gyroscopes and accelerometers. However, height drift caused by barometer is still unacceptable in a long-term duration. So a new method is proposed under the assumption that a person's height only changes on stairs in the building. Based on this assumption, the whole course of walking in a building is partitioned into floor phase and stair phase, according to a built HMM model. Height estimation can be corrected by considering apriori knowledge of buildings, and will be performed following different rules in the two phases. For the floor phase, the height remains consistent on the same floor, and minor drifts are suppressed inspired by Heuristic Drift Elimination (HDE). For the stair phase, after sufficient stair-related information is acquired, a Maximum Likelihood Estimation (MLE) method is used to estimate the height of each stair step, and drift-free height difference between two floors can be estimated using the estimated stair step height. The experimental results of the proposed approach demonstrate the effectiveness of the approach for height error suppression in SAR use with forced ventilation and sudden change in air pressure.

Ground based synthetic aperture radar (SAR) interferometry has already demonstrated its effectiveness in detecting manmade structures deformation and slow moving objects. A compact stepped frequency continuous wave SAR system for differential interferometry has been developed by the authors. In this paper the radar system characteristics are described together with the experimental results of displacements measurements.

A time domain 3D Rank Deficit Robust Capon Filter Bank (RD-RCFB) is presented. An additional preprocess and postprocess are adopted to transform the time domain model into the frequency domain model. And matrix vectorization is used to reduce the dimension. Optimal implement of the algorithm is discussed by comparing the cascading 1D RD-RCFB, cascading 2D RD-RCFB and 3D RD-RCFB. The algorithm outperforms the 3D Adaptive Sidelobe Reduction (ASR) and the 3D Amplitude and Phase Estimation of a Sinusoid(APES). Simulated planar aperture 3D image processing verified the algorithm.

This letter presents a novel change detection technique based on stepped-frequency continuous-wave synthetic aperture radar interferometry to detect human beings in motion inside a building. The proposed approach to moving-target indication consists of radar image formation, noncoherent energy change detection, and interferometric phase detection. Experimental results demonstrated that the proposed approach can dramatically reduce false alarms compared with a conventional noncoherent change detection approach.

Existing road extraction approaches have limitations on detecting roads covered by foliage, even for low-frequency SAR images. In this paper, a new approach is proposed to extract foliage-covered roads from low-frequency SAR images, by detecting road appurtenance alternatively. Road appurtenances alongside roads generally have consistent scattering and geometrical characteristics, which can be exploited by feature selection, classification and geometrical discrimination algorithms. The final result based on detection of guard trees indicates the effectiveness of the proposed approach.

For original paper see ibid., vol.7, no.6, p.634-42 (1999). We have no problem with the comments presented by Mathew and Kaimal (see ibid., p.415, 2005) regarding the investigation discussed in our original paper. The fact is that the general result in the investigated cases, that for a prisoner in both situations (noncooperative and selfish) the optional strategy is still to "confess" irrespective of the strategic choice made by the opponent party, really is in agreement with our original discussions. We thank the authors of the comments for their interest in our work, which has been recently expanded into "software testing via fuzzy strategic games" as well as "fighting terrorism in cyberspace using fuzzy strategic games." It is interesting to note that, especially in these two applications, the comments made by Mathew and Kaimal may make a significant impact.

The basic operations of fuzzy sets, such as negation, intersection, and union, usually are computed by applying the one-complement, minimum, and maximum operators to the membership functions of fuzzy sets. However, different decision agents may have different perceptions for these fuzzy operations. In this article, the concept of parameterized fuzzy operators will be introduced. A parameter α will be used to represent the degree of softness. The variance of α captures the differences of decision agents' subjective attitudes and characteristics, which result in their differing perceptions. The defined parameterized fuzzy operators also should satisfy the axiomatic requirements for the traditional fuzzy operators. A learning algorithm will be proposed to obtain the parameter α given a set of training data for each agent. In this article, the proposed parameterized fuzzy operators will be used in individual decision-making problems. An example is given to show the concept and application of the parameterized fuzzy operators. © 2003 Wiley Periodicals, Inc.

A game is a decision-making situation with many players, each having objectives that conflict with each other. The players involved in the game usually make their decisions under conditions of risk or uncertainty. In the paper, a fuzzy approach is proposed to solve the strategic game problem in which the pure strategy set for each player is already defined. Based on the concepts of fuzzy set theory, the approach uses a multicriteria decision-making method to obtain the optimal strategy in the game, a method which shows more advantages than the classical game methods. Moreover, with this approach, some useful conclusions are reached concerning the famous "prisoner's dilemma" problem in game theory.