Showing papers presented at "IEEE Asia-Pacific Conference on Synthetic Aperture Radar in 2015"

SAR ATR based on dividing CNN into CAE and SNN

Abstract: As for the problem of too long training time of convolution neural network (CNN), this paper proposes a fast training method for CNN in SAR automatic target recognition (ATR). The CNN is divided into two parts: one that contains all the convolution layers and sub-sampling layers is considered as convolutional auto-encoder (CAE) for unsupervised training to extract high-level features; the other that contains fully connected layers is regarded as shallow neural network (SNN) to work as a classifier. The experiment based on MSATR database shows that the proposed method can tremendously reduce the training time with little loss of recognition rate.

Target classification performance as a function of measurement uncertainty

Abstract: In this paper, we demonstrate target classification using the proposed features in previously reported research under measurement uncertainty conditions. The MSTAR dataset is widely used real target measurements in automatic target recognition society. Extremely high classification results of the dataset, which are over 90% correct classification, have been reported from some literatures. However, this high classification results could be acquired not only by the classification system, but also the cleanness of the dataset. Therefore, in this paper, more realistic target classification scenarios including target aspect angle estimation error, strong white Gaussian noise, and different combination of test and training targets are applied for classification and its corresponding results are examined. The proposed target feature extraction techniques show the robustness of the measurement uncertainties and excellent classification results.

Multi-static MIMO-SAR three dimensional deformation measurement system

Abstract: Herein is proposed a novel high-steep rock slope landslide monitoring system based on the configuration of multi-static multiple-input multiple-output synthetic aperture radar (MIMO-SAR). In this system, three spastically distributed radars are employed. Each is a Ku-band MIMO-SAR, transmitting mutually orthogonal waveforms and sensing the slope movement from different view angles. The three radars can then achieve the deform measurements in the three line of sight (LOS) directions. After an effective fusion of these measurements, the deform vector, with direction and amount, can be resolved. This vector is particularly important for risk warning as well as for understanding the motion mechanism of landslide. Moreover, the data acquisition period can be largely reduced, since the use of MIMO radar allows a simultaneous sampling of the scene echo. The time decorrelation is therefore very small, and hence high measurement accuracy can be expected. It can be proved that, using this system, the three-dimensional accuracy can reach 0.1 ∼ 1 mm. Simulations and experimental datasets are used to validate this proposal.

Ambiguities and image quality in staggered SAR

Abstract: Staggered SAR is an innovative synthetic aperture radar (SAR) concept, where the pulse repetition interval (PRI) is continuously varied. This, together with digital beamforming (DBF) in elevation, allows high-resolution imaging of a wide continuous swath without the need for a long antenna with multiple azimuth apertures. As an additional benefit, the energy of range and azimuth ambiguities is spread over large areas: Ambiguities therefore appear in the image as a noise-like disturbance rather than localized artifacts. An analytical expression for the range-ambiguity-to-signal ratio (RASR) in staggered SAR is provided and a novel method for the estimation of the azimuth ambiguity-to-signal ratio (AASR) is proposed. A C-band design example based on a planar antenna is shown as well. The impact of staggered SAR operation on image quality is further assessed using highly oversampled F-SAR airborne data.

A novel ship-detection technique for Sentinel-1 SAR data

Abstract: The paper shows a novel algorithm for ship-detection in Synthetic Aperture Radar (SAR) imagery. The algorithm is divided in three main steps: land mask rejection, detection and discrimination. In the first step land pixels are rejected by using Shuttle Radar Topography Mission (SRTM) data; in the second stage the potential ships are detected on a method based on the Generalized Likelihood Ratio Test (GLRT) and, finally, some targets are rejected by removing the azimuth ambiguities and by gathering the target pixels in clusters. The algorithm is tested on a Sentinel-1 image acquired over the Portsmouth harbour and compared with the outcomes coming from a Constant False Alarm Algorithm (CFAR).

Bistatic SAR imaging processing and experiment results using BeiDou-2/Compass-2 as illuminator of opportunity and a fixed receiver

Abstract: Global Navigation Satellite System (GNSS) provides opportunities for passive radar applications. This paper reports progress in space-surface bistatic synthetic aperture radar (SS-BiSAR) using Beidou-2 as transmitters of opportunity and a fixed ground-based receiver. Experiment is described and the obtained SAR image is presented and interpreted in detail. Results show that BiSAR using beidou-2 as illuminator is able to reflect different types of targets and SS-BiSAR using GNSS as illuminator of opportunity is potential to realize local area monitoring.

Sentinel-1 IWS mode support in the GAMMA software

Abstract: First results using the new Sentinel-1 SAR look very promising, but the special interferometrie wide-swath (IWS) data acquired in the FOPS mode makes InSAR processing challenging. Fhe steep azimuth spectra ramp in each burst results in very stringent co-registration requirements. Combining the data of the individual bursts and sub-swaths into consistent mosaics requires careful "bookkeeping" in the handling of the data and meta data and the large file sizes and high data throughputs require also a good performance. Considering these challenges good support from software is getting increasingly important. In this contribution we describe the Sentinel-1 support in the GAMMA Software, a high-level software package used by researchers, service providers and operational users in their SAR, InSAR and PSI work.

Development of short-range ground surveillance radar for moving target detection

Abstract: In this paper, we proposed a short-range ground surveillance FMCW (Frequency Modulated Continuous Wave) radar. For this work, we employed a 2D-FFT algorithm to detect moving targets, and a windowing scheme to reduce the distortion generated by frequency transition. Moreover, we designed a moving target finding method to extract moving targets from the remaining distortions. We developed a 24GHz FMCW radar system comprised of front-end and back-end modules. A newly designed algorithm was implemented in a micro-processor of the back-end module. When we tested the function of the developed radar system in the field, the range and velocity profiles of a moving car occupying the detection zone were extracted without distortion.

Multi-scale target detection in SAR image based on visual attention model

Abstract: This paper proposes a novel method for synthetic aperture radar (SAR) target detection by using multi-scale SAR images based on visual attention model, which can automatically find the vehicle targets from the complicated background with clutters such as trees and buildings. In our method, firstly, a saliency map is obtained from a Gaussian pyramid of the original SAR image, where the image scales are selected based on the prior size information of the targets to be detected in the image. Secondly, we use the method based on shifts of the focus of attention (FOA) in the saliency map to get a binary image. Finally, the clustering algorithm based on the prior length of targets is employed to extract the target candidate chips in the binary image. In the experiment based on the real SAR image, we compare the proposed method with the classical constant false alarm rate (CFAR) target detection method, which indicates that our method can detect vehicle targets in the image more quickly and with fewer false alarms.

Superpixel-based PolSAR images change detection

Abstract: Previous polarimetric synthetic aperture radar (PolSAR) images change detection methods are generally undertaken in the pixel scale, resulting in overlooking the semantic information. To solve this problem, this paper presents a superpixel-based PolSAR images change detection methods. Different from some previous methods, an improved SLIC superpixel segmentation method is introduced in polarimetric interferometric SAR (Pol-InSAR) images processing, which can segment two PolSAR images simultaneously. Moreover, along with the difference maps (DMs) generated based on the complex wishart distribution and the total intensity, we also use color information to generate the DM between PolSAR images, which has long been overlooked in PolSAR images change detection. Based on the segmentation results and three DMs, two different change detection schemes are proposed. One is to perform the majority voting to three change detection results using the above mentioned DMs. Another is to fuse the three DMs and generate a refined DM, and the change detection is performed on the refined DM. Two Radarsat-2 images acquired over Suzhou city, China, are used in our experiments to validate the proposed methods, and the experimental results show that the proposed schemes can improve the change detection results significantly.

Forest change detection from Sentinel-1 and ALOS-2 satellite images

Abstract: This research presents the performance of the recently launched C-band Sentinel-1 and L-band ALOS-2 satellite missions for detecting forest clear-cuts and thinning in the winter season in coniferous and deciduous forests within Estonia. SAR data were acquired from September 2014 to April 2015 to analyze the cross-polarized backscatter dynamics over 76 stands before and after the logging event. Both clear-cuts and thinning treatments were considered with stand size ranging from 0.11–6.44 ha and compared to the in situ reference data. The results show that the accuracy of logging detection is influenced by the weather conditions as well as the size of the stand. Both Sentinel-1 and ALOS-2 provide reliable forest clear-cut monitoring and are suitable for an operational change detection service for forest stands larger than 0.5 ha.

VISAR: A next generation interferometric radar for venus exploration

Abstract: Magellan, a NASA mission to Venus in the early 1990s, mapped nearly the entire surface of Venus with an S-band (12 cm) synthetic aperture radar and microwave radiometer and made radar altimeter measurements of the topography, [1]. These measurements revolutionized our understanding of the geomorphology, geology and geophysical processes that have shaped the evolution of the surface of Venus. However, the lack of finer resolution imagery and topography of the surface than that obtained by the Magellan mission has hampered the definitive answer to key questions concerning the processes and evolution of the surface of Venus. The Venus Emissivity, Radio Science, InSAR Topography And Spectroscopy (VERITAS) Mission is a proposed mission to Venus designed to obtain high resolution imagery and topography of the surface using an X-band radar configured as a single pass radar interferometer (called VISAR) coupled with a multispectral NIR emissivity mapping sensor.

Image formation method for spaceborne video SAR

Abstract: Spaceborne Video SAR imaging mode is a new imaging mode of SAR (Synthetic Aperture Radar), it is different from the traditional imaging mode. Operating in this mode, SAR provides continuous surveillance over a region of interest. Sequence images are processed afterwards and the information of the region is reappeared using video formed by sequence images. Obviously, this will bring many advantages for detecting and tracking moving-targets over the region of interest. Therefore, in this paper, the concept of spaceborne Video SAR is introduced at first, then the performances of spaceborne Video SAR, such as data process methodology, resolution and frame rate are discussed. Based on the above analysis, a modified Deramp Chirp Scaling (DCS) algorithm with an extra phase compensation factor is proposed to deal with spaceborne Video SAR data.

Two types of waveguide comparator for wideband monopulse antenna array application

Abstract: This paper presents two types of wideband Magic-Tee comparator for monopulse array application. They act as the key components for the design of wideband and low-profile monopulse comparator network. Both of them are easy to be integrated into the monopulse system and achieve good performance. Results and comparison of these two type comparators are also shown in the paper.

Bistatic PolInSAR for forest height estimation: Results from TanDEM-X

Abstract: The research work presents the initial results from bistatic dual-PolInSAR TerraSAR-X/TanDEM-X data. These results are the first reported observations of forest height derived using TerraSAR-X/TanDEM-X bistatic data over Indian tropical forests. The minimization approach with three stage inversion technique is applied on the data for forest height estimation. The results show potential of X-band SAR data for forest parameter estimation. Underestimation of height in forests due to low penetration capability of X-band and overestimation of height in agricultural areas are presented in the study.

A cognitive signals reconstruction algorithm based on compressed sensing

Abstract: Compressed Sensing (CS) theory has been widely used in radar signal processing field, and the reconstruction algorithm is the key to whether the original signal can be reconstructed from limited observations. However, the existing reconstruction algorithms either don't consider and remove the noise in signal reconstruction, or need the iterative estimation of noise variance during the signal reconstruction processing, which will lead the poor anti-noise performance or large computation load. In this paper, a cognitive signals reconstruction algorithm based on compressed sensing is proposed. In the method, the noise variance can be estimated by subspace decomposition method, and then the estimated noise variance is used as priori information in reconstruction algorithms to improve the reconstruction accuracy or reduce the computation load. As a result, the reconstruction algorithm performance can be improved effectively. Some simulation results illustrate the effectiveness of the proposed method.

Study on geo-location of sliding spotlight mode of GF-3 satellite

Abstract: GF-3 is the Chinese Synthetic Aperture Radar (SAR) satellite mission with scientific and commercial applications, which will be launched in 2016. It allows operation in Stripmap, ScanSAR and Sliding Spotlight modes in different polarizations. The Sliding Spotlight Mode is a new imaging mode widely used in high resolution SAR imaging. Considering the imaging processing model, we have modified the equivalent parameters on the base of the Range-Doppler model to improve the geometric accuracy for the sliding spotlight mode of GF-3. The simulation and TerraSAR-X experimental results show the the validity of the method.

Multi-band multi-mode SDR radar platform

Abstract: This paper presents the design results of the multi-band, multi-mode software-defined radar (SDR) system. The SDR platform consists of a multi-band RF modules of S, X, K-bands, and a multi-mode digital modules with a waveform generator for CW, Pulse, FMCW, and LFM Chirp as well as reconfigurable SDR-GUI software module for user interface. This platform can be used for various applications such as security monitoring, collision avoidance, traffic monitoring, and a radar imaging.

24-GHz microstrip grid array antenna for automotive radars application

Abstract: A single-ended microstrip grid array antenna has been designed and measured for 24 GHz automotive radars application. It has been designed on Rogers 5880 substrate with dimensions of 60×60×0.787 mm. The antenna exhibits 2.45 GHz impedance bandwidth, 19.26 dBi peak realized gain and broadside pencil-beam radiation pattern with low side lobe and weak cross-polarization.

Fusion of RADARSAT-2 imagery with LANDSAT-8 multispectral data for improving land cover classification performance using SVM

Abstract: Study of the land cover classification using multi-source data are very important for eco-environment monitoring, land use planning and climatic change detection. In this study, the utility of multi-source RADARSAT-2 and LANDSAT-8 multi-spectral images for improving land cover classification performance using Support Vector Machine (SVM) classifier. HH polarized C band RADARSAT-2 images were fused with the three band (6, 5, and 4) LANDSAT-8 multispectral image for land cover classification. Wavelet-based fusion (WT) techniques are implemented in the data fusion process. The Radial Basic Function (RBF) kernel function were used for SVM classifier in order to classify land cover types in the study area. The results of the SVM classification were compared with those using standard method Maximum Likelihood (ML) classifier, and it demonstrates a higher accuracy. Finally, it was indicated by the study that the fusion of SAR and optical images can significantly improve the classification accuracy with respect to use single dataset, and the SVM classifier could clearly outperform the standard method the ML classifier.

FastGBSAR case studies in China: Monitoring of a dam and instable slope

Abstract: Monitoring ground deformation and structural movements is an important task in order to prevent hazardous and harmful situations for the population. MetaSensing's FastGBSAR (Fast Ground-Based Synthetic Aperture Radar) is a novel sensor system, which allows the monitoring of deformation and vibration of natural and man-made structures. This paper describes two examples of application of the FastGBSAR to different sites in China providing examples of its usage for monitoring a dam, and an instable slope.

Application of microwave imaging in regional deformation monitoring using ground based SAR

Abstract: Regional deformation monitoring is important for early warning of deformation disasters. This paper focuses on the applications of microwave imaging in deformation monitoring based on ground-based SAR. Specifically, the paper mainly studies the microwave image reconstruction and deformation information extraction firstly. Then, matching principle and method of auxiliary terrain data and microwave image is analyzed. Based on the key technologies above, the paper carry out point target deformation monitoring experiment in laboratory and slope deformation monitoring experiment in a mine slope, which can verify the correctness of the models and the accuracy of the methods.

Discrete Stepped Strip (DI2S) for multi-swath acquisitions

Abstract: The Discrete Stepped Strip (DI2S) technique introduces an innovative method to use a SAR in time sharing improving the system capability and flexibility. In this paper, the approach used by DI2S Multi-Swath technique will be described highlighting the main advantages in terms of performance and application.

The linear fitting Kalman filter for nonlinear tracking

Abstract: Dynamic estimation in various synthetic aperture radar (SAR) applications often involves nonlinear models. The first-order Taylor approximation is usually used to linearize the nonlinear functions to perform estimation efficiently. However, the error generated by the first-order Taylor approximation cannot be negligible when dynamic systems are high nonlinearity or with large input errors. This paper proposes a linearization method by minimizing the difference between the nonlinear function and its linear approximation, and yields a better linear fitting function. A Kalman filter based on this principle is suggested. Simulation tests are conducted among the proposed linear fitting Kalman filter (LFKF), the extended Kalman filter (EKF) and the unscented Kalman filter (UKF). The results show that the LFKF obviously outperforms the EKF which uses the first-order Taylor approximation, and the LFKF can achieve similar estimation accuracy of the UKF with less computational cost.

S- and X-band SAR data fusion

Abstract: This paper investigates the benefits deriving from introducing a wavelet-transform-based fusion framework for multi-frequency Synthetic Aperture Radar (SAR) data. A specific application is considered in the assessment of the fused classification map derived and this is the discrimination of different kinds of oil in sea. S-band and X-band datasets, concurrently acquired from the same airborne platform, have here been used. The findings suggest that fusing S-band and X-band SAR data does improve the oil type discrimination between crude oil and diesel oil used in the exercise, although a more quantitative analysis should be conducted in the future to measure the degree of improvement.

Glacier surface velocity estimation using stokes vector correlation

Abstract: Cryosphere plays a crucial role in regulating local and global climate. Glaciers form an important component of this frozen part of the Earth's system. They exist over a prolonged period and are largest reservoir of freshwater on Earth. The meltwater from the glaciers during warmer seasons contribute to the river systems in absence of other sources. The run-off is also useful for agriculture, power generation, and is rich in alluvium. Retreating glaciers gives rise to pro-glacial lakes formed by damming action of moraine or ice. Rupturing of ice dams have caused serious damage to infrastructure and human lives in the past. Such useful and dynamic characteristics of a glacier motivate us to study its movement. Monitoring glaciers through in-situ field measurements is a cumbersome process. Over the past decade, glaciers have been repeatedly observed through microwave sensors on-board various satellites. Various techniques have been proposed in the literature to estimate glacier velocity using microwave observations. Over the recent past, the trend in glacier velocity monitoring has shifted from interferometric tracking to intensity tracking. In this paper, we propose a method based on Stokes vector correlation of time lapse microwave observations. This method is proposed as an improvement over the conventional intensity correlation technique.

The DLR SAR calibration center

Abstract: A SAR system has to be calibrated for quantitative measurement. During this process all essential parameters of a SAR image are linked to their geophysical quantity. This includes the location of the image pixel, its backscattering characteristics (in amplitude and in phase) and polarimetric information. The Microwave and Radar Institute of the DLR has gained a lot of experience in these calibration procedures during the last decades and has developed special methods and dedicated reference targets for spaceborne SAR system calibration. In example of calibration results obtained for different spaceborne SAR systems the DLR SAR Calibration Center is presented.

A dual-band FMCW radar for through-wall detection

Abstract: In this paper, a dual-band, frequency-modulated continuous-wave (FMCW) radar for short-range through-wall detection is proposed and implemented. This radar adopts a shared-aperture antenna technique for reducing antenna area and high-speed chirping to avoid flicker noise. It operates at the S-band (3 GHz) and the X-band (9 GHz), with 486 MHz chirp bandwidth and 860 MHz chirp bandwidth, respectively. The chirp rates are 11,050 GHz/s and 22,000 GHz/s at the S- and X-bands, respectively. The radar has successfully detected a target through a wooden wall.

Sidelobe level reduction in uniformly-fed E-plane coupled horn antenna array

Abstract: In this paper we deal with uniformly-fed E-plane coupled horn array antenna. We find a configuration to have a maximum gain, and a configuration to achieve sidelobe reduction. A combination of space taper, beamwidth taper and squint taper is applied. We present array geometries with 1–4 elements, and it is shown that sidelobe reduction of approximately 10 dB's can be obtained, while maintaining a reasonable value for the gain.

Data volume reduction in high-resolution wide-swath SAR systems

Abstract: High-resolution wide-swath (HRWS) synthetic aperture radar (SAR) systems are very attractive for the observation of dynamic processes on the Earth's surface, but they are also associated with a huge data volume. In order to comply with azimuth ambiguity requirements, in fact, a pulse repetition frequency (PRF) much higher than the required processed Doppler bandwidth (PBW) is often desirable. The data volume can be drastically reduced, if on-board Doppler filtering and decimation are performed prior to downlink. A finite impulse response (FIR) filter with a relatively small number of taps suffices to suppress the additional ambiguous components and recover the original impulse response. This strategy is also applicable and especially relevant to staggered SAR systems, where on-board Doppler filtering and resampling can be jointly implemented.