Synthetic Aperture Radar Ocean Images Research Articles

SAR Imaging Algorithm of Ocean Waves Based on Optimum Subaperture.

Synthetic Aperture Radar (SAR) is widely applied to the field of ocean remote sensing. Clear SAR images are the basis for ocean information acquisitions, such as parameter retrieval of ocean waves and wind field inversion of the ocean surface. However, the SAR ocean images are usually blurred, which seriously affects the acquisition of ocean information. The reasons for the wave blurring in SAR images mainly include the following two aspects. One is that when SAR observes the ocean, the motion of ocean waves will have a greater impact on imaging quality. The other is that the ocean’s surface is seriously decorrelated within the integration time. In order to obtain clear SAR images of ocean waves, a SAR imaging algorithm of ocean waves based on the optimum subaperture is proposed, aiming at the above two aspects. The optimum focus setting of the ocean waves is calculated, drawing support from the azimuth phase velocity of the dominant wave. The optimum subaperture is further calculated according to the proposed new evaluation, namely, F. Finally, according to the optimum focus setting and the optimum subaperture, the dominant wave is refocused, and a clear SAR image of the dominant wave can be obtained. The proposed algorithm was applied to airborne L-band and P-band SAR data. Furthermore, the proposed algorithm was compared with present methods, and the results sufficiently demonstrated the effectiveness and superiority of the proposed algorithm.

An accelerated nonlocal means algorithm for synthetic aperture radar ocean image despeckling

An accelerated nonlocal means algorithm for synthetic aperture radar ocean image despeckling

An Azimuth Antenna Pattern Estimation Method Based on Doppler Spectrum in SAR Ocean Images.

In synthetic aperture radar (SAR) ocean remote sensing, it is very difficult to estimate an accurate azimuth antenna pattern (AAP) from low-scattering SAR images without strong scattering targets. Therefore, an azimuth antenna pattern estimation method based on Doppler spectrum in SAR ocean images is proposed in this paper. In order to preserve the complete AAP information, an azimuth unweighted matched filter is used to re-image the SAR raw data in the proposed method. Then, the shape factor of AAP can be obtained by linear statistics of the relationship between Doppler center and edge frequency spectrum in Doppler spectrum of each distance gate. In addition, the impact of the uniformity and signal-to-noise ratio of SAR ocean images on the estimation results are also analyzed by simulation. Finally, the feasibility of proposed method is verified by data from ERS-2 (European remote sensing satellite (ERS) was the European Space Agency’s first Earth-observing satellite). Experimental results show that the AAP estimated by proposed method has a good estimation result.

A Fast Level Set Method for Synthetic Aperture Radar Ocean Image Segmentation

Segmentation of high noise imagery like Synthetic Aperture Radar (SAR) images is still one of the most challenging tasks in image processing. While level set, a novel approach based on the analysis of the motion of an interface, can be used to address this challenge, the cell-based iterations may make the process of image segmentation remarkably slow, especially for large-size images. For this reason fast level set algorithms such as narrow band and fast marching have been attempted. Built upon these, this paper presents an improved fast level set method for SAR ocean image segmentation. This competent method is dependent on both the intensity driven speed and curvature flow that result in a stable and smooth boundary. Notably, it is optimized to track moving interfaces for keeping up with the point-wise boundary propagation using a single list and a method of fast up-wind scheme iteration. The list facilitates efficient insertion and deletion of pixels on the propagation front. Meanwhile, the local up-wind scheme is used to update the motion of the curvature front instead of solving partial differential equations. Experiments have been carried out on extraction of surface slick features from ERS-2 SAR images to substantiate the efficacy of the proposed fast level set method.

Internal wave detection and location in SAR images using wavelet transform

An efficient computational framework far the extraction of mesoscale features, i.e., internal waves, present in synthetic aperture radar (SAR) images is discussed. The manifestation of internal wave packets on SAR ocean images has always been of considerable interest to oceanographers. If, as seems likely, the gray tone patterns of these images can be confirmed to correspond to trough and crest patterns of internal waves, a great deal can be learned about internal waves from satellite data. The first problem in a mesoscale detection system is to distinguish sea from land in the SAR imagery. A method for coastline detection based on a sequence of basic processing procedures followed by a contour tracing algorithm is introduced to obtain sea-land separation to enhance the internal wave detection problem. The utility of wavelet analysis as a tool for automatic oceanic internal wave detection and orientation from SAR images is then examined using the two-dimensional (2D) wavelet transform based on the multiscale gradient detection method. The authors show that the evolution of local maxima of the wavelet transform across scales characterize the local shape of these quasilinear periodic structures. The results from this study, on several ERS-1 and RADARSAT SAR images, show that wavelet analysis is an excellent tool for detecting and locating internal wave features from satellite images against internal wave look-alikes.

Focusing of synthetic aperture radar ocean images with long integration times

Synthetic aperture radar (SAR) images obtained in the SAR and X Band Ocean Nonlinearities: Chesapeake Light Tower (SAXON:CLT) experiment are processed with long integration times (6 s) and analyzed to study the effects of focusing. Two images with near‐azimuth‐traveling waves were chosen for the study. The first image consists of relatively short wavelength wind waves traveling in the same general direction as the aircraft. The second image consists of a long Atlantic swell traveling in the opposite direction to the aircraft. At these long integration times the image spectral intensities are found to be sensitive to the focus setting. The spectral intensity at the optimum focus is 400% of that at zero focus for the first image and 167% for the second image. The focusing curves for both images agree well with those predicted by a model developed by several groups and referred to here as the “consensus” model. This model predicts an optimum focus setting that is equal to one half of the effective phase speed of the dominant wave in the azimuth direction. The velocity bunching model underpredicts the optimum focus setting significantly. The study concludes that in long‐integration‐time SAR processing of surface waves, such as the spotlight mode, the image contrast is sensitively dependent on the focus setting and that the optimum focus setting is given by one half of the effective phase speed of the dominant surface wave.

Azimuth smearing in synthetic aperture radar ocean image spectra from the Norwegian Continental Shelf Experiment of 1988

Along‐track resolution loss in synthetic aperture radar ocean images is investigated using airborne C band data collected at different altitudes. A method is developed for qualifying and quantifying the amount of smearing. The smearing can be described as a low‐pass‐filtering process acting on the image spectra even in a nonlinear region of imaging. The dependency on the ratio of range to platform velocity and the radar incidence angle is presented. An intrinsic value for the subresolution velocity spread (i.e., the rms facet velocity) is estimated to be in the region σu = 0.59–0.73 m/s for the measurements available on March 11, 1988, with significant wave height ≈ 4 m and wind speed ≈ 8 m/s. These values for σu allow a conversion of the smearing to ERS 1 geometry giving a spatial smearing σx in the range 68–84 m which corresponds to a resolution varying between 300 and 370 m in the along‐track direction.

The consequences of sampling variability on the estimation of wave number and propagation direction from spaceborne SAR image spectra

Even if it were true that SAR (synthetic aperture radar) ocean surface imagery provides a perfect noiseless representation of ocean surface wave height, each spectrum computed from such imagery would be but a single realization of the ensemble-mean ocean spectrum. There would be sampling variability associated with the parameters of dominant wave number and propagation direction extracted from such a spectrum. The present study addresses two questions: (i) what statistical distribution is applicable to the spectra of SAR ocean images? and (ii) what are the consequences of such statistics on the precision with which wave number and propagation direction can be extracted? An examination is made of spectra computed from ocean imagery acquired during the SIR-B (Shuttle Imaging Radar-B) mission. >

Comparisons of simulated and actual synthetic aperture radar gravity wave images

The synthetic aperture radar (SAR) images obtained with the Jet Propulsion Laboratory's L band system in the Tower Ocean Wave and Radar Dependence (TOWARD) oceanographic experiment are compared with images generated by a simulation program implementing a SAR ocean imaging model based on two‐scale hydrodynamic and electromagnetic scattering approximate models. A critically regarded test of a theory of SAR ocean imaging is its prediction of the best image focus dependence on surface motion. The primary means of comparison here is an estimate of the best focus parameter that uses a subimage cross‐correlation technique. The focus parameter estimates for both the actual and simulated images show (1) a reversal in sign with reversal in dominant longwave direction relative to the SAR direction, (2) a magnitude increase with an increase in magnitude of the angle between the dominent long wave and SAR axes, and (3) an independence of the altitude and also the range‐to‐velocity ratio. All equivalent velocity estimates are of the order of the dominant longwave phase velocity, normalized by the SAR vehicle velocity, and agree well qualitatively though there are some quantitative differences. These behaviors are in agreement with a related analytical prediction. The “visibility” of the long waves' SAR image artifacts, for this TOWARD data set, increased with increased altitude and also with increased range‐to‐velocity ratio for both the actual and simulated images, as was predicted by a related analysis. Agreement of spectral density estimates of the actual and simulated SAR images generally required an order of magnitude increase in the strength of the hydrodynamic interaction of the long and short waves. This may indicate, as have surface measurements in TOWARD and other experiments, that the two‐scale model does not fully describe this nonlinear interaction. With this interaction adjustment, a comparison of actual and simulated images using an alternative focus criterion showed very close agreement in some cases.

Theory of synthetic aperture radar ocean imaging: A MARSEN view

This paper reviews basic synthetic aperture radar (SAR) theory of ocean wave imaging mechanisms, using both known work and recent experimental and theoretical results from the Marine Remote Sensing (MARSEN) Experiment. Several viewpoints that have contributed to the field are drawn together in a general analysis of the backscatter statistics of a moving sea surface. A common focus for different scattering models is provided by the mean image impulse response function, which is shown to be identical to the (spatially varying) frequency variance spectrum of the local complex reflectivity coefficient. From the analysis has emerged a more complete view of the SAR imaging phenomenon than has been previously available. A new, generalized imaging model is proposed.

Improvement in the estimation of dominate wavenumber and direction from spaceborne SAR image spectra when corrected for ocean surface movement

Because a synthetic aperture radar (SAR) is a Doppler device, ocean surface movement can degrade a SAR ocean image. Ocean surface movement causes the redistribution of apparent radar cross section in a SAR image to neighboring image elements. It is demonstrated in this paper that this aspect of SAR imaging can be treated as a simple resolution loss in a linear system. Such an approach explains the falloff in response at high azimuth wavenumbers experienced in SAR image spectra. More importantly, this approach allows for a straightforward method to alleviate the problem of ocean surface motion. Correction of SAR image spectra using a linear systems approach results in SAR image spectra in which the locations in wavenumber-space of surface wave spectral peaks are more consistent with the estimated location of storm sources than are the locations of peaks from uncorrected spectra.

Synthetic aperture radar imaging of ocean waves during the marineland experiment

X - and L -band simultaneously obtained synthetic aperture radar (SAR) data of ocean gravity waves collected during the Marineland Experiment were analyzed using wave contrast measurements. The Marineland data collected in 1975 represents a unique historical data set for testing still-evolving theoretical models of the SAR ocean wave imaging process. The wave contrast measurements referred to are direct measurements of the backscatter variation between wave crests and troughs. These modulation depth measurements, which are indicators of wave detectability, were made as a function of: a) the settings used in processing the SAR signal histories to partially account for wave motion; b) wave propagation direction with respect to radar look direction for both X - and L -band SAR data; c) SAR resolution; and d) number of coherent looks. The contrast measurements indicated that ocean waves imaged by a SAR are most discernible when X -band frequency is used (as compared to L -band), and when the ocean waves are traveling in the range direction. Ocean waves can be detected by both X - and L -band SAR, provided that the radar surface resolution is small compared to the ocean wavelength (at least 1/4 of the ocean wavelength is indicated by this work). Finally, wave detection with L -band SAR can be improved by adjusting the focal distance and rotation of the cylindrical telescope in the SAR optical processor to account for wave motion. The latter adjustments are found to be proportional to a value that is near the wave phase velocity.