Zenghui Zhang

Edge Extraction for Polarimetric SAR Images Using Degenerate Filter With Weighted Maximum Likelihood Estimation

The classic region-based filter for edge extraction for polarimetric synthetic aperture radar images is theoretically founded and efficient. However, in practical use, its performance is limited because the assumption of independence and identical distribution is often not met, particularly in heterogeneous areas. In this letter, we present a degenerate filter design integrated with the weighted maximum likelihood estimation to overcome this limitation. The performance of the proposed methodology is presented and analyzed on both simulated and real experimental data sets using visual presentation, as well as numerical evaluation and comparison with the classic method. They both demonstrate the availability and advantage of the proposed method.

Representation and Spatially Adaptive Segmentation for PolSAR Images Based on Wedgelet Analysis

It is believed that it is essential to take the spatial adaptivity into the segmentation method for polarimetric synthetic aperture radar (PolSAR) images. The size and shape of each segment and the strength of the relationship of neighboring pixels need to depend on the local spatial complexity of the scene. The wedgelet framework provides a promising analysis tool for spatial information. The major advantage of the wedgelet analysis is that it captures the geometrical structure of images at multiple scales, with the local spatial complexity taken into consideration. Hence, in this paper, we propose a wedgelet approximation and analysis framework specially designed for PolSAR data. Based on this framework, a spatially adaptive representation and segmentation method is constructed and presented. It mainly consists of three parts: first, the multiscale wedgelet decomposition is applied to the PolSAR image, and the local geometrical information is captured in an optimal way; then, the image is segmented in a spatially adaptive manner by the multiscale wedgelet representation in the form of the regularized optimization, which keeps a balance between the approximation and parsimony of the representation; the final part is the spatial-complexity-adaptive segmentation refinement based on the Wishart Markov random field model. The performance of the proposed method is presented and analyzed on two experimental data sets, with visual presentation and numerical evaluation. It is also compared with an existing and theoretically well-founded segmentation method. The experiments and results demonstrate the availability and advantage of the proposed method.

Technical Evaluation of Sentinel-1 IW Mode Cross-Pol Radar Backscattering from the Ocean Surface in Moderate Wind Condition

The Sentinel-1 synthetic aperture radar (SAR) allows sufficient resources for cross-pol wind speed retrievals over the ocean. In this paper, we present technical evaluation on wind retrieval from both Sentinel-1A and Sentinel-1B IW cross-pol images. Algorithms are based on the existing theoretical and empirical ones derived from the RADARSAT-2 cross-pol data. First, to better understand the Sentinel-1 observed normalized radar cross section (NRCS) values under various environmental conditions, we constructed a dataset that integrates SAR images with wind field information from scatterometer measurements. There are 11,883 matchup data in the experimental dataset. We then calculated the systemic noise floor of Sentinel-1 IW mode, and presented its unique noise characteristics among different sub-bands. Based on the calculated NESZ measurements, the noise is removed for all matchup data. Empirical relationships among the noise free NRCS σ VH 0 , wind speed, wind direction, and radar incidence angle are analyzed for each sub-band, and a piecewise model is proposed. We showed that a larger correlation coefficient, r, is achieved by including both wind direction and incidence terms in the model. Validation against scatterometer measurements showed the suitability of the proposed model.

A Space-Time Coding Scheme With Time and Frequency Comb-Like Chirp Waveforms for MIMO-SAR

A new space-time coding (STC) scheme for multiple-input multiple-output synthetic aperture radar systems is proposed in this paper. In the new scheme, two successive signal periods are put into one transmit duration to minimize the time-variant channel effect. The even and odd components of the transmitted waveforms are modulated into distinct Doppler frequencies in the azimuth direction and can be separated by bandpass filter in the range-Doppler domain. The presented STC scheme can suppress the interchannel ambiguous energy caused by the time-variant channel responses, which typically occur in conventional STC schemes, and increase the signal-to-noise ratio of the echoes by a decoding matrix. To further suppress the interference, a waveform called time and frequency comb-like chirp (TFCC) is proposed. With the application of the complementary comb-like structure in the time and frequency domains, two TFCC waveforms are short-term shift-orthogonal and constant envelope. Moreover, these corresponding TFCC waveforms can share the same antenna and time gate without sacrificing any TFCC waveforms’ peak level. The theoretical analysis and simulation results illustrate the feasibility of the proposed scheme.

Convolutional Neural Network for SAR image classification at patch level

Convolutional Neural Network (CNN) has attracted much attention for feature learning and image classification, mostly related to close range photography. As a benchmark work, we trained a relatively large CNN to classify SAR image patches into five different categories, where the image patches tiled and annotated from a typical TerraSAR-X spotlight scene of Wuhan, China. The neural network designed in this paper consists of seven layers, including one input layer, two convolutional layers where each followed by a max-pooling layer, as well as two fully-connected layers with a final five-class softmax. Using the toolkit caffe, we achieved the training and testing accuracy of 85.7% and 85.6% respectively, which is considerably better than the traditional feature extraction and classification based SVM method and shows great potential of CNN used for SAR image interpretation. In order to accelerate the training process, a very efficient GPU implementation was employed.

Tropical Cyclone Boundary Layer Rolls in Synthetic Aperture Radar Imagery

Marine atmospheric boundary layer (MABL) roll plays an important role in the turbulent exchange of momentum, sensible heat, and moisture throughout MABL of tropical cyclone (TC). Hence, rolls are believed to be closely related to TC’s development, intensification, and decay processes. Spaceborne synthetic aperture radar (SAR) provides a unique capability to image the sea surface imprints of quasi-linear streaks induced by the MABL rolls within a TC. In this study, sixteen SAR images, including three images acquired during three major hurricanes: Irma, Jose, and Maria in the 2017 Atlantic hurricane season, were utilized to systematically map the distribution and wavelength of MABL rolls under the wide range of TC intensities. The images were acquired by SAR onboard RADARSAT-1/2, ENVISAT, and SENTINEL-1 satellites. Our findings are in agreement with the previous one case study of Hurricane Katrina (2005), showing the roll wavelengths are between 600 and 1,600 m. We also find that there exist roll imprints in eyewall and rainbands, although the boundary layer heights are shallower there. Besides, the spatial distribution of roll wavelengths is asymmetrical. The roll wavelengths are found to be the shortest around the storm center, increase and then decrease with distance from storm center, reaching the peak values in the range of d 22d , where d is defined as the physical location to TC centers normalized by the radius of maximum wind. These MABL roll characteristics cannot be derived using conventional aircraft and land-based Doppler radar observations.

Preliminary exploration of SAR image land cover classification with noisy labels

Synthetic Aperture Radar (SAR) image land cover classification is an important task in SAR image interpretation. Supervised learning, such as Convolutional Neural Network (CNN), demands instances which are accurately labeled. However, a large amount of accurately labeled SAR images are difficult to produce. In this paper, a Probability Transition CNN (PTCNN) is proposed for patch-level SAR image land cover classification with noisy labels. Firstly, deep features are extracted by a CNN model, followed by a probabilistic transition model, where true labels are treated as hidden variables and the posterior probabilities of true labels are transferred into their noisy versions. The whole network is trained with Caffe in a uniform fashion and a land cover database is used to produce noisy labels, which are randomly chosen with various proportions. Experimental results demonstrate that the proposed PTCNN model is robust to noise and gives a promising classification performance. Therefore, the PTCNN model may lower the standards for the quality of image labels, and shows its availability in practical applications.

Preliminary evaluation of vessel detectability for Sentinel-1 SAR data

Performance of ship detection is influenced by synthetic aperture radar (SAR) imaging characteristics and environmental conditions. In this paper, aiming at evaluating vessel detectability for Sentinel-1 SAR data, a model based on a large-scale Sentinel-1A vessel chips database is established. The model sensitivity is analyzed by simulation data. In the experiment, by inputting the parameters of imaging characteristics (incidence angle, polarization, spatial resolution) and environmental conditions (wind speed, wind direction, sea state) of a specific Sentinel-1 image, the minimum detectable vessel length can be estimated. Further validations demonstrate the availability of the estimated minimum detectable vessel length.

Circulate shifted OFDM chirp waveform diversity design with digital beamforming for MIMO SAR

Waveform diversity design has always been the key to multiple-input multiple-output (MIMO) synthetic aperture radar (SAR) systems, and it is known that synthetic integral side lobe ratio (SISLR) is a more optimal indicator than the integral side lobe ratio (ISLR) to evaluate the orthogonality between different MIMO SAR waveforms. This paper presents proof that it is difficult to obtain the SISLR of an existing waveform such that it is sufficiently low to achieve high SNR for SAR imaging. Thus, it is necessary to find a way to separate these MIMO SAR waveforms in other domains, for example, a spatial domain by digital beamforming (DBF). Learning from Krieger’s idea of short-term shift-orthogonal waveforms and using joint time-frequency transforms (Gabor transform) to prove that for most existing SAR waveforms cyclic shift is a good operation with which to generate short-term shift-orthogonal waveforms. This paper presents the designs of four circulate shifted OFDM chirp waveforms, which have a much lower SISLR and retain all the advantages of the classical chirp waveform such as a large time-bandwidth product, constant amplitude, implementation simplicity, and good Doppler tolerance.

OpenSARShip 2.0: A large-volume dataset for deeper interpretation of ship targets in Sentinel-1 imagery

In the era of big data for synthetic aperture radar (SAR), multiple SAR systems have come into service, including Sentinel-1. The support of large-volume datasets is the key to the deep interpretation of ship targets in Sentinel-1 imagery. In this paper, we present OpenSARShip 2.0, the newest and upgraded version of OpenSARShip, which is dedicated to the deeper interpretation of SAR imagery for marine surveillance. Based on the higher requirements for ship target interpretation, the OpenSARShip 2.0, covering 34528 SAR ship chips with automatic identification system (AIS) information, possesses three improvements: large volume, interference labeling, and type levels, which are described in this paper. In addition, an application case involving the geometric parameter extraction is introduced to demonstrate the applicability of the large dataset.